diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5545 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.07462338510330675,
+  "eval_steps": 500,
+  "global_step": 500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "clip_ratio": 0.0,
+      "completion_length": 838.0089645385742,
+      "epoch": 0.0001492467702066135,
+      "grad_norm": 0.031563565135002136,
+      "kl": 0.0,
+      "learning_rate": 3.3333333333333335e-07,
+      "loss": 0.0,
+      "num_tokens": 153987.0,
+      "reward": 0.31626437886734493,
+      "reward_std": 0.22882110241334885,
+      "rewards/code_reward": 0.3127822265960276,
+      "rewards/format_reward": 0.034821428707800806,
+      "step": 1
+    },
+    {
+      "clip_ratio": 0.0,
+      "epoch": 0.000298493540413227,
+      "grad_norm": 0.03154602274298668,
+      "kl": 0.0,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 0.0,
+      "step": 2
+    },
+    {
+      "clip_ratio": 0.0010276091998093762,
+      "epoch": 0.00044774031061984047,
+      "grad_norm": 0.030962640419602394,
+      "kl": 0.00018510222434997559,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.0002,
+      "step": 3
+    },
+    {
+      "clip_ratio": 0.0010089212846651208,
+      "completion_length": 782.9911155700684,
+      "epoch": 0.000596987080826454,
+      "grad_norm": 0.03505287691950798,
+      "kl": 0.00018844008445739746,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": -0.0,
+      "num_tokens": 310016.0,
+      "reward": 0.38136595563264564,
+      "reward_std": 0.2664957612287253,
+      "rewards/code_reward": 0.3755623744800687,
+      "rewards/format_reward": 0.05803571408614516,
+      "step": 4
+    },
+    {
+      "clip_ratio": 0.0011241662487009307,
+      "epoch": 0.0007462338510330675,
+      "grad_norm": 0.03434380516409874,
+      "kl": 0.00018966197967529297,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.0001,
+      "step": 5
+    },
+    {
+      "clip_ratio": 0.0009702896441012854,
+      "epoch": 0.0008954806212396809,
+      "grad_norm": 0.033386170864105225,
+      "kl": 0.0001881122589111328,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": -0.0,
+      "step": 6
+    },
+    {
+      "clip_ratio": 0.0009430443981273129,
+      "completion_length": 886.3482513427734,
+      "epoch": 0.0010447273914462945,
+      "grad_norm": 0.03664536774158478,
+      "kl": 0.0001901388168334961,
+      "learning_rate": 2.3333333333333336e-06,
+      "loss": -0.0002,
+      "num_tokens": 471944.0,
+      "reward": 0.2568418255104916,
+      "reward_std": 0.24231004383182153,
+      "rewards/code_reward": 0.24987753484310815,
+      "rewards/format_reward": 0.06964285799767822,
+      "step": 7
+    },
+    {
+      "clip_ratio": 0.001033175552947796,
+      "epoch": 0.001193974161652908,
+      "grad_norm": 0.032914530485868454,
+      "kl": 0.0001964867115020752,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": -0.0,
+      "step": 8
+    },
+    {
+      "clip_ratio": 0.0009873282078842749,
+      "epoch": 0.0013432209318595214,
+      "grad_norm": 0.03524191677570343,
+      "kl": 0.00020045042037963867,
+      "learning_rate": 3e-06,
+      "loss": 0.0,
+      "step": 9
+    },
+    {
+      "clip_ratio": 0.0012342520167294424,
+      "completion_length": 792.6607513427734,
+      "epoch": 0.001492467702066135,
+      "grad_norm": 0.03604082018136978,
+      "kl": 0.00021690130233764648,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": -0.0001,
+      "num_tokens": 632563.0,
+      "reward": 0.3381975887459703,
+      "reward_std": 0.2540945124346763,
+      "rewards/code_reward": 0.33221544651314616,
+      "rewards/format_reward": 0.05982143012806773,
+      "step": 10
+    },
+    {
+      "clip_ratio": 0.0011053664220526116,
+      "epoch": 0.0016417144722727485,
+      "grad_norm": 0.0362468957901001,
+      "kl": 0.00021219253540039062,
+      "learning_rate": 3.6666666666666666e-06,
+      "loss": -0.0001,
+      "step": 11
+    },
+    {
+      "clip_ratio": 0.0011532589833223028,
+      "epoch": 0.0017909612424793619,
+      "grad_norm": 0.03517252579331398,
+      "kl": 0.00022751092910766602,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.0,
+      "step": 12
+    },
+    {
+      "clip_ratio": 0.0008506373051204719,
+      "completion_length": 906.1696853637695,
+      "epoch": 0.0019402080126859754,
+      "grad_norm": 0.04052340239286423,
+      "kl": 0.000179976224899292,
+      "learning_rate": 4.333333333333334e-06,
+      "loss": -0.0,
+      "num_tokens": 806035.0,
+      "reward": 0.16382310001790756,
+      "reward_std": 0.20915061386767775,
+      "rewards/code_reward": 0.16087666610837914,
+      "rewards/format_reward": 0.029464285646099597,
+      "step": 13
+    },
+    {
+      "clip_ratio": 0.0009509306000836659,
+      "epoch": 0.002089454782892589,
+      "grad_norm": 0.032756347209215164,
+      "kl": 0.0001971125602722168,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": -0.0,
+      "step": 14
+    },
+    {
+      "clip_ratio": 0.001035670431519975,
+      "epoch": 0.0022387015530992023,
+      "grad_norm": 0.04692872241139412,
+      "kl": 0.00018927454948425293,
+      "learning_rate": 5e-06,
+      "loss": -0.0,
+      "step": 15
+    },
+    {
+      "clip_ratio": 0.0007350842770392774,
+      "completion_length": 834.8929023742676,
+      "epoch": 0.002387948323305816,
+      "grad_norm": 0.031541600823402405,
+      "kl": 0.0002028346061706543,
+      "learning_rate": 4.999952797253148e-06,
+      "loss": -0.0,
+      "num_tokens": 962081.0,
+      "reward": 0.3517821817804361,
+      "reward_std": 0.17664615268586203,
+      "rewards/code_reward": 0.3481214540079236,
+      "rewards/format_reward": 0.03660714376019314,
+      "step": 16
+    },
+    {
+      "clip_ratio": 0.0006742932200722862,
+      "epoch": 0.0025371950935124295,
+      "grad_norm": 0.030405128374695778,
+      "kl": 0.00020629167556762695,
+      "learning_rate": 4.9998111909931225e-06,
+      "loss": 0.0,
+      "step": 17
+    },
+    {
+      "clip_ratio": 0.0006983668281463906,
+      "epoch": 0.002686441863719043,
+      "grad_norm": 0.03497137501835823,
+      "kl": 0.0002194046974182129,
+      "learning_rate": 4.999575187161439e-06,
+      "loss": -0.0002,
+      "step": 18
+    },
+    {
+      "clip_ratio": 0.0012954147914570058,
+      "completion_length": 804.232177734375,
+      "epoch": 0.0028356886339256566,
+      "grad_norm": 0.037687718868255615,
+      "kl": 0.00023552775382995605,
+      "learning_rate": 4.9992447956603455e-06,
+      "loss": 0.0,
+      "num_tokens": 1125333.0,
+      "reward": 0.30283497480559163,
+      "reward_std": 0.29476781317498535,
+      "rewards/code_reward": 0.2947099723969586,
+      "rewards/format_reward": 0.08125000062864274,
+      "step": 19
+    },
+    {
+      "clip_ratio": 0.0011280627804808319,
+      "epoch": 0.00298493540413227,
+      "grad_norm": 0.03715205565094948,
+      "kl": 0.0002586245536804199,
+      "learning_rate": 4.998820030352409e-06,
+      "loss": -0.0001,
+      "step": 20
+    },
+    {
+      "clip_ratio": 0.001265888700800133,
+      "epoch": 0.0031341821743388833,
+      "grad_norm": 0.03643012046813965,
+      "kl": 0.0002815723419189453,
+      "learning_rate": 4.998300909059929e-06,
+      "loss": -0.0001,
+      "step": 21
+    },
+    {
+      "clip_ratio": 0.0012057735548296478,
+      "completion_length": 837.4375419616699,
+      "epoch": 0.003283428944545497,
+      "grad_norm": 0.04015280678868294,
+      "kl": 0.00030815601348876953,
+      "learning_rate": 4.997687453564198e-06,
+      "loss": -0.0,
+      "num_tokens": 1279382.0,
+      "reward": 0.29113965947180986,
+      "reward_std": 0.2496117369737476,
+      "rewards/code_reward": 0.2873003660934046,
+      "rewards/format_reward": 0.03839285811409354,
+      "step": 22
+    },
+    {
+      "clip_ratio": 0.001202213983560796,
+      "epoch": 0.0034326757147521104,
+      "grad_norm": 0.03490012139081955,
+      "kl": 0.00031620264053344727,
+      "learning_rate": 4.9969796896045775e-06,
+      "loss": -0.0001,
+      "step": 23
+    },
+    {
+      "clip_ratio": 0.0011879785979544977,
+      "epoch": 0.0035819224849587238,
+      "grad_norm": 0.03548375889658928,
+      "kl": 0.0003438591957092285,
+      "learning_rate": 4.996177646877426e-06,
+      "loss": -0.0,
+      "step": 24
+    },
+    {
+      "clip_ratio": 0.0008819098402454983,
+      "completion_length": 791.4643211364746,
+      "epoch": 0.0037311692551653375,
+      "grad_norm": 0.037079837173223495,
+      "kl": 0.00047838687896728516,
+      "learning_rate": 4.995281359034851e-06,
+      "loss": 0.0001,
+      "num_tokens": 1425573.0,
+      "reward": 0.17420367311569862,
+      "reward_std": 0.1679191511357203,
+      "rewards/code_reward": 0.16581081453477964,
+      "rewards/format_reward": 0.08392857149010524,
+      "step": 25
+    },
+    {
+      "clip_ratio": 0.0008276363059849245,
+      "epoch": 0.003880416025371951,
+      "grad_norm": 0.03505370393395424,
+      "kl": 0.0005742311477661133,
+      "learning_rate": 4.994290863683296e-06,
+      "loss": 0.0,
+      "step": 26
+    },
+    {
+      "clip_ratio": 0.0009231225321855163,
+      "epoch": 0.004029662795578565,
+      "grad_norm": 0.03283383697271347,
+      "kl": 0.0006587505340576172,
+      "learning_rate": 4.99320620238196e-06,
+      "loss": 0.0,
+      "step": 27
+    },
+    {
+      "clip_ratio": 0.0013180975292925723,
+      "completion_length": 812.6250381469727,
+      "epoch": 0.004178909565785178,
+      "grad_norm": 0.04038141295313835,
+      "kl": 0.0008317828178405762,
+      "learning_rate": 4.99202742064106e-06,
+      "loss": 0.0001,
+      "num_tokens": 1579160.0,
+      "reward": 0.4461335694650188,
+      "reward_std": 0.2936822731862776,
+      "rewards/code_reward": 0.4360442637989763,
+      "rewards/format_reward": 0.10089285846333951,
+      "step": 28
+    },
+    {
+      "clip_ratio": 0.001309220588154858,
+      "epoch": 0.004328156335991791,
+      "grad_norm": 0.034344371408224106,
+      "kl": 0.0011445283889770508,
+      "learning_rate": 4.990754567919917e-06,
+      "loss": -0.0001,
+      "step": 29
+    },
+    {
+      "clip_ratio": 0.00168330199085176,
+      "epoch": 0.004477403106198405,
+      "grad_norm": 0.037127457559108734,
+      "kl": 0.0016373395919799805,
+      "learning_rate": 4.989387697624881e-06,
+      "loss": -0.0,
+      "step": 30
+    },
+    {
+      "clip_ratio": 0.0007538784138887422,
+      "completion_length": 852.9196853637695,
+      "epoch": 0.004626649876405018,
+      "grad_norm": 0.03424752503633499,
+      "kl": 0.0019218921661376953,
+      "learning_rate": 4.987926867107095e-06,
+      "loss": 0.0,
+      "num_tokens": 1741264.0,
+      "reward": 0.2446170602052007,
+      "reward_std": 0.1605516797862947,
+      "rewards/code_reward": 0.23408133629709482,
+      "rewards/format_reward": 0.10535714169964194,
+      "step": 31
+    },
+    {
+      "clip_ratio": 0.0009339907319372287,
+      "epoch": 0.004775896646611632,
+      "grad_norm": 0.03297942504286766,
+      "kl": 0.002615690231323242,
+      "learning_rate": 4.986372137660078e-06,
+      "loss": 0.0001,
+      "step": 32
+    },
+    {
+      "clip_ratio": 0.0010387174697825685,
+      "epoch": 0.004925143416818246,
+      "grad_norm": 0.03285828232765198,
+      "kl": 0.0036895275115966797,
+      "learning_rate": 4.984723574517165e-06,
+      "loss": -0.0,
+      "step": 33
+    },
+    {
+      "clip_ratio": 0.0010631408213157556,
+      "completion_length": 897.4286155700684,
+      "epoch": 0.005074390187024859,
+      "grad_norm": 0.03713543340563774,
+      "kl": 0.0033243894577026367,
+      "learning_rate": 4.9829812468487655e-06,
+      "loss": 0.0002,
+      "num_tokens": 1908976.0,
+      "reward": 0.3391252807341516,
+      "reward_std": 0.29133173869922757,
+      "rewards/code_reward": 0.32323241684935056,
+      "rewards/format_reward": 0.15892857313156128,
+      "step": 34
+    },
+    {
+      "clip_ratio": 0.001270327098609414,
+      "epoch": 0.005223636957231472,
+      "grad_norm": 0.04065223038196564,
+      "kl": 0.004559874534606934,
+      "learning_rate": 4.981145227759457e-06,
+      "loss": 0.0001,
+      "step": 35
+    },
+    {
+      "clip_ratio": 0.0015276970298145898,
+      "epoch": 0.005372883727438086,
+      "grad_norm": 0.037328362464904785,
+      "kl": 0.006010532379150391,
+      "learning_rate": 4.979215594284924e-06,
+      "loss": 0.0001,
+      "step": 36
+    },
+    {
+      "clip_ratio": 0.0008515377248841105,
+      "completion_length": 948.6161041259766,
+      "epoch": 0.005522130497644699,
+      "grad_norm": 0.0387113131582737,
+      "kl": 0.004285603761672974,
+      "learning_rate": 4.977192427388722e-06,
+      "loss": 0.0,
+      "num_tokens": 2095140.0,
+      "reward": 0.34807550528785214,
+      "reward_std": 0.2688880347413942,
+      "rewards/code_reward": 0.33495048189070076,
+      "rewards/format_reward": 0.1312500003259629,
+      "step": 37
+    },
+    {
+      "clip_ratio": 0.000987556606560247,
+      "epoch": 0.005671377267851313,
+      "grad_norm": 0.03865530341863632,
+      "kl": 0.004810154438018799,
+      "learning_rate": 4.9750758119588824e-06,
+      "loss": 0.0002,
+      "step": 38
+    },
+    {
+      "clip_ratio": 0.0009448816053918563,
+      "epoch": 0.0058206240380579265,
+      "grad_norm": 0.041891518980264664,
+      "kl": 0.005879461765289307,
+      "learning_rate": 4.972865836804349e-06,
+      "loss": 0.0001,
+      "step": 39
+    },
+    {
+      "clip_ratio": 0.0010522838042561489,
+      "completion_length": 944.3303852081299,
+      "epoch": 0.00596987080826454,
+      "grad_norm": 0.039550624787807465,
+      "kl": 0.008145570755004883,
+      "learning_rate": 4.970562594651254e-06,
+      "loss": -0.0,
+      "num_tokens": 2274125.0,
+      "reward": 0.3804548046900891,
+      "reward_std": 0.19460752099985257,
+      "rewards/code_reward": 0.36295478582178475,
+      "rewards/format_reward": 0.17499999917345122,
+      "step": 40
+    },
+    {
+      "clip_ratio": 0.0010110271978192031,
+      "epoch": 0.006119117578471153,
+      "grad_norm": 0.040130678564310074,
+      "kl": 0.008675038814544678,
+      "learning_rate": 4.968166182139026e-06,
+      "loss": -0.0,
+      "step": 41
+    },
+    {
+      "clip_ratio": 0.0008353690973308403,
+      "epoch": 0.006268364348677767,
+      "grad_norm": 0.04057696834206581,
+      "kl": 0.009891927242279053,
+      "learning_rate": 4.9656766998163306e-06,
+      "loss": -0.0001,
+      "step": 42
+    },
+    {
+      "clip_ratio": 0.0012194093724247068,
+      "completion_length": 896.2678871154785,
+      "epoch": 0.006417611118884381,
+      "grad_norm": 0.038462087512016296,
+      "kl": 0.009711623191833496,
+      "learning_rate": 4.963094252136865e-06,
+      "loss": 0.0001,
+      "num_tokens": 2438179.0,
+      "reward": 0.37517865585687105,
+      "reward_std": 0.25164606305770576,
+      "rewards/code_reward": 0.3574107780586928,
+      "rewards/format_reward": 0.1776785693364218,
+      "step": 43
+    },
+    {
+      "clip_ratio": 0.0012110178904549684,
+      "epoch": 0.006566857889090994,
+      "grad_norm": 0.03634024038910866,
+      "kl": 0.010460138320922852,
+      "learning_rate": 4.960418947454958e-06,
+      "loss": 0.0001,
+      "step": 44
+    },
+    {
+      "clip_ratio": 0.001134861959144473,
+      "epoch": 0.0067161046592976075,
+      "grad_norm": 0.04107284173369408,
+      "kl": 0.01217663288116455,
+      "learning_rate": 4.957650898021038e-06,
+      "loss": 0.0002,
+      "step": 45
+    },
+    {
+      "clip_ratio": 0.0010649148716765922,
+      "completion_length": 788.4018230438232,
+      "epoch": 0.006865351429504221,
+      "grad_norm": 0.0408879816532135,
+      "kl": 0.016703248023986816,
+      "learning_rate": 4.954790219976915e-06,
+      "loss": 0.0003,
+      "num_tokens": 2592497.0,
+      "reward": 0.17120053875260055,
+      "reward_std": 0.1869892734248424,
+      "rewards/code_reward": 0.15093267243355513,
+      "rewards/format_reward": 0.20267857098951936,
+      "step": 46
+    },
+    {
+      "clip_ratio": 0.0011155054035043577,
+      "epoch": 0.007014598199710834,
+      "grad_norm": 0.05487113445997238,
+      "kl": 0.018317341804504395,
+      "learning_rate": 4.95183703335091e-06,
+      "loss": 0.0002,
+      "step": 47
+    },
+    {
+      "clip_ratio": 0.0010340367061871802,
+      "epoch": 0.0071638449699174475,
+      "grad_norm": 0.04240145906805992,
+      "kl": 0.021048426628112793,
+      "learning_rate": 4.948791462052819e-06,
+      "loss": 0.0001,
+      "step": 48
+    },
+    {
+      "clip_ratio": 0.0011539331317180768,
+      "completion_length": 805.6339588165283,
+      "epoch": 0.007313091740124062,
+      "grad_norm": 0.0386325865983963,
+      "kl": 0.017290830612182617,
+      "learning_rate": 4.945653633868716e-06,
+      "loss": 0.0003,
+      "num_tokens": 2741962.0,
+      "reward": 0.4239194593974389,
+      "reward_std": 0.24492623138939962,
+      "rewards/code_reward": 0.40409802555222996,
+      "rewards/format_reward": 0.1982142855413258,
+      "step": 49
+    },
+    {
+      "clip_ratio": 0.0009078447328647599,
+      "epoch": 0.007462338510330675,
+      "grad_norm": 0.03812170773744583,
+      "kl": 0.016597270965576172,
+      "learning_rate": 4.942423680455584e-06,
+      "loss": 0.0002,
+      "step": 50
+    },
+    {
+      "clip_ratio": 0.0010736608801380498,
+      "epoch": 0.007611585280537288,
+      "grad_norm": 0.03711409866809845,
+      "kl": 0.015314102172851562,
+      "learning_rate": 4.939101737335802e-06,
+      "loss": 0.0001,
+      "step": 51
+    },
+    {
+      "clip_ratio": 0.0010047486684925389,
+      "completion_length": 914.5714664459229,
+      "epoch": 0.007760832050743902,
+      "grad_norm": 0.038698915392160416,
+      "kl": 0.015361785888671875,
+      "learning_rate": 4.935687943891447e-06,
+      "loss": 0.0002,
+      "num_tokens": 2909543.0,
+      "reward": 0.43717449135147035,
+      "reward_std": 0.15000206953845918,
+      "rewards/code_reward": 0.41494232891272986,
+      "rewards/format_reward": 0.22232142696157098,
+      "step": 52
+    },
+    {
+      "clip_ratio": 0.0009450064771954203,
+      "epoch": 0.007910078820950515,
+      "grad_norm": 0.03570646047592163,
+      "kl": 0.017791748046875,
+      "learning_rate": 4.932182443358458e-06,
+      "loss": 0.0002,
+      "step": 53
+    },
+    {
+      "clip_ratio": 0.000910184779058909,
+      "epoch": 0.00805932559115713,
+      "grad_norm": 0.03823816031217575,
+      "kl": 0.0196990966796875,
+      "learning_rate": 4.928585382820616e-06,
+      "loss": 0.0001,
+      "step": 54
+    },
+    {
+      "clip_ratio": 0.0012172754632047145,
+      "completion_length": 815.2500438690186,
+      "epoch": 0.008208572361363742,
+      "grad_norm": 0.0404551737010479,
+      "kl": 0.02109670639038086,
+      "learning_rate": 4.924896913203376e-06,
+      "loss": 0.0003,
+      "num_tokens": 3056830.0,
+      "reward": 0.4185403302544728,
+      "reward_std": 0.25729377527022734,
+      "rewards/code_reward": 0.38871889375150204,
+      "rewards/format_reward": 0.2982142888940871,
+      "step": 55
+    },
+    {
+      "clip_ratio": 0.001118028167184093,
+      "epoch": 0.008357819131570356,
+      "grad_norm": 0.041043564677238464,
+      "kl": 0.021697044372558594,
+      "learning_rate": 4.921117189267535e-06,
+      "loss": 0.0003,
+      "step": 56
+    },
+    {
+      "clip_ratio": 0.000911217556677002,
+      "epoch": 0.00850706590177697,
+      "grad_norm": 0.040621694177389145,
+      "kl": 0.02174830436706543,
+      "learning_rate": 4.917246369602742e-06,
+      "loss": 0.0002,
+      "step": 57
+    },
+    {
+      "clip_ratio": 0.0009711457878438523,
+      "completion_length": 783.9732513427734,
+      "epoch": 0.008656312671983583,
+      "grad_norm": 0.04503508657217026,
+      "kl": 0.015195846557617188,
+      "learning_rate": 4.9132846166208355e-06,
+      "loss": 0.0002,
+      "num_tokens": 3209882.0,
+      "reward": 0.2656380059197545,
+      "reward_std": 0.2678962650243193,
+      "rewards/code_reward": 0.23849514660832938,
+      "rewards/format_reward": 0.27142857061699033,
+      "step": 58
+    },
+    {
+      "clip_ratio": 0.0008419685891567497,
+      "epoch": 0.008805559442190197,
+      "grad_norm": 0.04335927218198776,
+      "kl": 0.015669822692871094,
+      "learning_rate": 4.9092320965490365e-06,
+      "loss": 0.0001,
+      "step": 59
+    },
+    {
+      "clip_ratio": 0.0008221774314733921,
+      "epoch": 0.00895480621239681,
+      "grad_norm": 0.04567364230751991,
+      "kl": 0.016145706176757812,
+      "learning_rate": 4.905088979422971e-06,
+      "loss": 0.0001,
+      "step": 60
+    },
+    {
+      "clip_ratio": 0.0007586929878016235,
+      "completion_length": 986.6250419616699,
+      "epoch": 0.009104052982603424,
+      "grad_norm": 0.03794924542307854,
+      "kl": 0.012533783912658691,
+      "learning_rate": 4.900855439079536e-06,
+      "loss": 0.0001,
+      "num_tokens": 3396775.0,
+      "reward": 0.2646394136536401,
+      "reward_std": 0.17509737799991854,
+      "rewards/code_reward": 0.24338940205052495,
+      "rewards/format_reward": 0.21250000200234354,
+      "step": 61
+    },
+    {
+      "clip_ratio": 0.0008951123400038341,
+      "epoch": 0.009253299752810036,
+      "grad_norm": 0.0516357384622097,
+      "kl": 0.013607978820800781,
+      "learning_rate": 4.8965316531496055e-06,
+      "loss": 0.0001,
+      "step": 62
+    },
+    {
+      "clip_ratio": 0.001083579034457216,
+      "epoch": 0.00940254652301665,
+      "grad_norm": 0.040371041744947433,
+      "kl": 0.015568733215332031,
+      "learning_rate": 4.892117803050578e-06,
+      "loss": 0.0001,
+      "step": 63
+    },
+    {
+      "clip_ratio": 0.0008599141901868279,
+      "completion_length": 867.6964645385742,
+      "epoch": 0.009551793293223265,
+      "grad_norm": 0.044309817254543304,
+      "kl": 0.023859024047851562,
+      "learning_rate": 4.887614073978761e-06,
+      "loss": 0.0004,
+      "num_tokens": 3552029.0,
+      "reward": 0.28711679979460314,
+      "reward_std": 0.2155274103861302,
+      "rewards/code_reward": 0.25979537097737193,
+      "rewards/format_reward": 0.27321428805589676,
+      "step": 64
+    },
+    {
+      "clip_ratio": 0.000977037544544146,
+      "epoch": 0.009701040063429877,
+      "grad_norm": 0.043455298990011215,
+      "kl": 0.02324676513671875,
+      "learning_rate": 4.883020654901609e-06,
+      "loss": 0.0004,
+      "step": 65
+    },
+    {
+      "clip_ratio": 0.0009009606010295101,
+      "epoch": 0.009850286833636491,
+      "grad_norm": 0.04464162513613701,
+      "kl": 0.02041912078857422,
+      "learning_rate": 4.878337738549785e-06,
+      "loss": 0.0002,
+      "step": 66
+    },
+    {
+      "clip_ratio": 0.0010641216322255787,
+      "completion_length": 672.2232532501221,
+      "epoch": 0.009999533603843104,
+      "grad_norm": 0.0937013179063797,
+      "kl": 0.019114971160888672,
+      "learning_rate": 4.873565521409082e-06,
+      "loss": 0.0003,
+      "num_tokens": 3686797.0,
+      "reward": 0.5510467411950231,
+      "reward_std": 0.13944192277267575,
+      "rewards/code_reward": 0.5136360162869096,
+      "rewards/format_reward": 0.37410714593715966,
+      "step": 67
+    },
+    {
+      "clip_ratio": 0.000909985425096238,
+      "epoch": 0.010148780374049718,
+      "grad_norm": 0.040670786052942276,
+      "kl": 0.018751859664916992,
+      "learning_rate": 4.868704203712173e-06,
+      "loss": 0.0003,
+      "step": 68
+    },
+    {
+      "clip_ratio": 0.0008242175144914654,
+      "epoch": 0.010298027144256332,
+      "grad_norm": 0.04120355844497681,
+      "kl": 0.018866777420043945,
+      "learning_rate": 4.86375398943021e-06,
+      "loss": 0.0001,
+      "step": 69
+    },
+    {
+      "clip_ratio": 0.0009311497433373006,
+      "completion_length": 895.2143325805664,
+      "epoch": 0.010447273914462945,
+      "grad_norm": 0.0390135832130909,
+      "kl": 0.01908421516418457,
+      "learning_rate": 4.858715086264274e-06,
+      "loss": 0.0002,
+      "num_tokens": 3850404.0,
+      "reward": 0.26960674731526524,
+      "reward_std": 0.17457441904116422,
+      "rewards/code_reward": 0.23549959086813033,
+      "rewards/format_reward": 0.3410714268684387,
+      "step": 70
+    },
+    {
+      "clip_ratio": 0.0009219656994901015,
+      "epoch": 0.010596520684669559,
+      "grad_norm": 0.041038449853658676,
+      "kl": 0.01819467544555664,
+      "learning_rate": 4.853587705636646e-06,
+      "loss": 0.0002,
+      "step": 71
+    },
+    {
+      "clip_ratio": 0.0008795039802862448,
+      "epoch": 0.010745767454876171,
+      "grad_norm": 0.03821442276239395,
+      "kl": 0.017704248428344727,
+      "learning_rate": 4.84837206268195e-06,
+      "loss": 0.0001,
+      "step": 72
+    },
+    {
+      "clip_ratio": 0.0012245200068718987,
+      "completion_length": 813.1250343322754,
+      "epoch": 0.010895014225082785,
+      "grad_norm": 0.03698420152068138,
+      "kl": 0.014069795608520508,
+      "learning_rate": 4.8430683762381195e-06,
+      "loss": 0.0003,
+      "num_tokens": 4010767.0,
+      "reward": 0.45059246465098113,
+      "reward_std": 0.2168489265604876,
+      "rewards/code_reward": 0.41996744694188237,
+      "rewards/format_reward": 0.306249993853271,
+      "step": 73
+    },
+    {
+      "clip_ratio": 0.0011484919050417375,
+      "epoch": 0.011044260995289398,
+      "grad_norm": 0.035879626870155334,
+      "kl": 0.012859582901000977,
+      "learning_rate": 4.837676868837213e-06,
+      "loss": 0.0002,
+      "step": 74
+    },
+    {
+      "clip_ratio": 0.0010865089680009987,
+      "epoch": 0.011193507765496012,
+      "grad_norm": 0.03598187118768692,
+      "kl": 0.013086557388305664,
+      "learning_rate": 4.832197766696085e-06,
+      "loss": 0.0002,
+      "step": 75
+    },
+    {
+      "clip_ratio": 0.0011177412579854717,
+      "completion_length": 851.6250400543213,
+      "epoch": 0.011342754535702626,
+      "grad_norm": 0.04002436622977257,
+      "kl": 0.027965068817138672,
+      "learning_rate": 4.826631299706887e-06,
+      "loss": 0.0005,
+      "num_tokens": 4166006.0,
+      "reward": 0.25119215465383604,
+      "reward_std": 0.19108223204966635,
+      "rewards/code_reward": 0.2239600099856034,
+      "rewards/format_reward": 0.27232143259607255,
+      "step": 76
+    },
+    {
+      "clip_ratio": 0.0009378863978781737,
+      "epoch": 0.011492001305909239,
+      "grad_norm": 0.03956011310219765,
+      "kl": 0.027362942695617676,
+      "learning_rate": 4.820977701427424e-06,
+      "loss": 0.0003,
+      "step": 77
+    },
+    {
+      "clip_ratio": 0.0010317562919226475,
+      "epoch": 0.011641248076115853,
+      "grad_norm": 0.040139809250831604,
+      "kl": 0.029706120491027832,
+      "learning_rate": 4.81523720907136e-06,
+      "loss": 0.0002,
+      "step": 78
+    },
+    {
+      "clip_ratio": 0.0010767242492875084,
+      "completion_length": 840.508955001831,
+      "epoch": 0.011790494846322466,
+      "grad_norm": 0.04374147206544876,
+      "kl": 0.013130903244018555,
+      "learning_rate": 4.809410063498254e-06,
+      "loss": -0.0001,
+      "num_tokens": 4338571.0,
+      "reward": 0.4022022826829925,
+      "reward_std": 0.25021205260418355,
+      "rewards/code_reward": 0.37372013460844755,
+      "rewards/format_reward": 0.28482142509892583,
+      "step": 79
+    },
+    {
+      "clip_ratio": 0.0010619452350510983,
+      "epoch": 0.01193974161652908,
+      "grad_norm": 0.042508602142333984,
+      "kl": 0.013546943664550781,
+      "learning_rate": 4.8034965092034656e-06,
+      "loss": 0.0001,
+      "step": 80
+    },
+    {
+      "clip_ratio": 0.0010448531356814783,
+      "epoch": 0.012088988386735694,
+      "grad_norm": 0.04649941623210907,
+      "kl": 0.01339411735534668,
+      "learning_rate": 4.797496794307889e-06,
+      "loss": -0.0,
+      "step": 81
+    },
+    {
+      "clip_ratio": 0.0010828481463249773,
+      "completion_length": 921.9821815490723,
+      "epoch": 0.012238235156942306,
+      "grad_norm": 0.04039299488067627,
+      "kl": 0.018650054931640625,
+      "learning_rate": 4.791411170547545e-06,
+      "loss": 0.0002,
+      "num_tokens": 4510139.0,
+      "reward": 0.3364243805408478,
+      "reward_std": 0.30479453783482313,
+      "rewards/code_reward": 0.31017436436377466,
+      "rewards/format_reward": 0.26250000204890966,
+      "step": 82
+    },
+    {
+      "clip_ratio": 0.0008476070252072532,
+      "epoch": 0.01238748192714892,
+      "grad_norm": 0.040362074971199036,
+      "kl": 0.01748943328857422,
+      "learning_rate": 4.785239893263017e-06,
+      "loss": 0.0001,
+      "step": 83
+    },
+    {
+      "clip_ratio": 0.001010585496260319,
+      "epoch": 0.012536728697355533,
+      "grad_norm": 0.039599236100912094,
+      "kl": 0.016507625579833984,
+      "learning_rate": 4.778983221388742e-06,
+      "loss": 0.0001,
+      "step": 84
+    },
+    {
+      "clip_ratio": 0.0010595632738841232,
+      "completion_length": 799.1607475280762,
+      "epoch": 0.012685975467562147,
+      "grad_norm": 0.041213374584913254,
+      "kl": 0.016191959381103516,
+      "learning_rate": 4.77264141744214e-06,
+      "loss": 0.0003,
+      "num_tokens": 4668714.0,
+      "reward": 0.4136724097770639,
+      "reward_std": 0.23826690285932273,
+      "rewards/code_reward": 0.38706526932219276,
+      "rewards/format_reward": 0.26607142575085163,
+      "step": 85
+    },
+    {
+      "clip_ratio": 0.0008205887170333881,
+      "epoch": 0.012835222237768762,
+      "grad_norm": 0.04150375723838806,
+      "kl": 0.015417337417602539,
+      "learning_rate": 4.766214747512603e-06,
+      "loss": 0.0002,
+      "step": 86
+    },
+    {
+      "clip_ratio": 0.0008396551329497015,
+      "epoch": 0.012984469007975374,
+      "grad_norm": 0.03998123109340668,
+      "kl": 0.01720881462097168,
+      "learning_rate": 4.759703481250331e-06,
+      "loss": 0.0001,
+      "step": 87
+    },
+    {
+      "clip_ratio": 0.0010611350435283384,
+      "completion_length": 836.7500419616699,
+      "epoch": 0.013133715778181988,
+      "grad_norm": 0.03714749217033386,
+      "kl": 0.0200653076171875,
+      "learning_rate": 4.753107891855015e-06,
+      "loss": 0.0001,
+      "num_tokens": 4824940.0,
+      "reward": 0.4083272517891601,
+      "reward_std": 0.22058491280768067,
+      "rewards/code_reward": 0.38091653838637285,
+      "rewards/format_reward": 0.2741071404889226,
+      "step": 88
+    },
+    {
+      "clip_ratio": 0.0009375711833854439,
+      "epoch": 0.0132829625483886,
+      "grad_norm": 0.0368531309068203,
+      "kl": 0.019817352294921875,
+      "learning_rate": 4.746428256064375e-06,
+      "loss": 0.0001,
+      "step": 89
+    },
+    {
+      "clip_ratio": 0.0010657701477612136,
+      "epoch": 0.013432209318595215,
+      "grad_norm": 0.047454558312892914,
+      "kl": 0.02184915542602539,
+      "learning_rate": 4.7396648541425534e-06,
+      "loss": 0.0,
+      "step": 90
+    },
+    {
+      "clip_ratio": 0.0010593721226541675,
+      "completion_length": 861.2053928375244,
+      "epoch": 0.013581456088801827,
+      "grad_norm": 0.03824659064412117,
+      "kl": 0.015380144119262695,
+      "learning_rate": 4.732817969868348e-06,
+      "loss": 0.0004,
+      "num_tokens": 4993061.0,
+      "reward": 0.4410945254203398,
+      "reward_std": 0.21284229948651046,
+      "rewards/code_reward": 0.40627308818511665,
+      "rewards/format_reward": 0.34821428428404033,
+      "step": 91
+    },
+    {
+      "clip_ratio": 0.0010110190596606117,
+      "epoch": 0.013730702859008442,
+      "grad_norm": 0.040716975927352905,
+      "kl": 0.015566825866699219,
+      "learning_rate": 4.7258878905233095e-06,
+      "loss": 0.0003,
+      "step": 92
+    },
+    {
+      "clip_ratio": 0.0010006381999119185,
+      "epoch": 0.013879949629215056,
+      "grad_norm": 0.03767255321145058,
+      "kl": 0.015938758850097656,
+      "learning_rate": 4.718874906879688e-06,
+      "loss": 0.0002,
+      "step": 93
+    },
+    {
+      "clip_ratio": 0.0009179275557471556,
+      "completion_length": 751.1339683532715,
+      "epoch": 0.014029196399421668,
+      "grad_norm": 0.042217355221509933,
+      "kl": 0.022792816162109375,
+      "learning_rate": 4.711779313188231e-06,
+      "loss": 0.0003,
+      "num_tokens": 5139257.0,
+      "reward": 0.40494063566438854,
+      "reward_std": 0.1955110931303352,
+      "rewards/code_reward": 0.3761906255967915,
+      "rewards/format_reward": 0.2875000007916242,
+      "step": 94
+    },
+    {
+      "clip_ratio": 0.0008862965933076339,
+      "epoch": 0.014178443169628283,
+      "grad_norm": 0.04299935698509216,
+      "kl": 0.0221099853515625,
+      "learning_rate": 4.70460140716584e-06,
+      "loss": 0.0003,
+      "step": 95
+    },
+    {
+      "clip_ratio": 0.0009494263367741951,
+      "epoch": 0.014327689939834895,
+      "grad_norm": 0.04077346622943878,
+      "kl": 0.021993637084960938,
+      "learning_rate": 4.697341489983076e-06,
+      "loss": 0.0002,
+      "step": 96
+    },
+    {
+      "clip_ratio": 0.0008807632484604255,
+      "completion_length": 809.6428909301758,
+      "epoch": 0.01447693671004151,
+      "grad_norm": 0.040974438190460205,
+      "kl": 0.02227783203125,
+      "learning_rate": 4.6899998662515215e-06,
+      "loss": 0.0003,
+      "num_tokens": 5283564.0,
+      "reward": 0.4401598667027429,
+      "reward_std": 0.3118339345091954,
+      "rewards/code_reward": 0.4056062815361656,
+      "rewards/format_reward": 0.34553571231663227,
+      "step": 97
+    },
+    {
+      "clip_ratio": 0.0010328342650609557,
+      "epoch": 0.014626183480248123,
+      "grad_norm": 0.04546511545777321,
+      "kl": 0.021175384521484375,
+      "learning_rate": 4.682576844011007e-06,
+      "loss": 0.0002,
+      "step": 98
+    },
+    {
+      "clip_ratio": 0.0010472134017618373,
+      "epoch": 0.014775430250454736,
+      "grad_norm": 0.04181186109781265,
+      "kl": 0.021422386169433594,
+      "learning_rate": 4.675072734716678e-06,
+      "loss": 0.0004,
+      "step": 99
+    },
+    {
+      "clip_ratio": 0.0009307468735642033,
+      "completion_length": 739.9553871154785,
+      "epoch": 0.01492467702066135,
+      "grad_norm": 0.04855009913444519,
+      "kl": 0.018518447875976562,
+      "learning_rate": 4.667487853225931e-06,
+      "loss": 0.0003,
+      "num_tokens": 5432463.0,
+      "reward": 0.4610493981745094,
+      "reward_std": 0.24288451514439657,
+      "rewards/code_reward": 0.4116743914783001,
+      "rewards/format_reward": 0.49375000409781933,
+      "step": 100
+    },
+    {
+      "clip_ratio": 0.001017617063553189,
+      "epoch": 0.015073923790867963,
+      "grad_norm": 0.0445452444255352,
+      "kl": 0.020605087280273438,
+      "learning_rate": 4.659822517785203e-06,
+      "loss": 0.0003,
+      "step": 101
+    },
+    {
+      "clip_ratio": 0.0009316829600720666,
+      "epoch": 0.015223170561074577,
+      "grad_norm": 0.04415549710392952,
+      "kl": 0.022787094116210938,
+      "learning_rate": 4.6520770500166165e-06,
+      "loss": 0.0002,
+      "step": 102
+    },
+    {
+      "clip_ratio": 0.0010779687218018807,
+      "completion_length": 861.6339664459229,
+      "epoch": 0.015372417331281191,
+      "grad_norm": 0.03548543527722359,
+      "kl": 0.02224445343017578,
+      "learning_rate": 4.644251774904487e-06,
+      "loss": 0.0003,
+      "num_tokens": 5597916.0,
+      "reward": 0.12403731828089803,
+      "reward_std": 0.16697289876174182,
+      "rewards/code_reward": 0.09421588902478106,
+      "rewards/format_reward": 0.2982142874971032,
+      "step": 103
+    },
+    {
+      "clip_ratio": 0.0009617375380912563,
+      "epoch": 0.015521664101487804,
+      "grad_norm": 0.036081403493881226,
+      "kl": 0.023712158203125,
+      "learning_rate": 4.636347020781684e-06,
+      "loss": 0.0003,
+      "step": 104
+    },
+    {
+      "clip_ratio": 0.0011486706807772862,
+      "epoch": 0.015670910871694418,
+      "grad_norm": 0.03618353605270386,
+      "kl": 0.024709224700927734,
+      "learning_rate": 4.6283631193158605e-06,
+      "loss": 0.0003,
+      "step": 105
+    },
+    {
+      "clip_ratio": 0.0011008622332155937,
+      "completion_length": 892.1518287658691,
+      "epoch": 0.01582015764190103,
+      "grad_norm": 0.04874338582158089,
+      "kl": 0.011252880096435547,
+      "learning_rate": 4.620300405495532e-06,
+      "loss": 0.0003,
+      "num_tokens": 5773913.0,
+      "reward": 0.32322888064663857,
+      "reward_std": 0.19707379283499904,
+      "rewards/code_reward": 0.28242531220894307,
+      "rewards/format_reward": 0.4080357113853097,
+      "step": 106
+    },
+    {
+      "clip_ratio": 0.0008924784815462772,
+      "epoch": 0.015969404412107643,
+      "grad_norm": 0.04063912108540535,
+      "kl": 0.01187896728515625,
+      "learning_rate": 4.612159217616022e-06,
+      "loss": 0.0001,
+      "step": 107
+    },
+    {
+      "clip_ratio": 0.0010843967938853893,
+      "epoch": 0.01611865118231426,
+      "grad_norm": 0.03972407057881355,
+      "kl": 0.012783527374267578,
+      "learning_rate": 4.603939897265268e-06,
+      "loss": 0.0001,
+      "step": 108
+    },
+    {
+      "clip_ratio": 0.0010632504654495278,
+      "completion_length": 760.1875267028809,
+      "epoch": 0.01626789795252087,
+      "grad_norm": 0.04420414939522743,
+      "kl": 0.03521156311035156,
+      "learning_rate": 4.595642789309492e-06,
+      "loss": 0.0004,
+      "num_tokens": 5917315.0,
+      "reward": 0.5145110483281314,
+      "reward_std": 0.2714524638140574,
+      "rewards/code_reward": 0.4761181781068444,
+      "rewards/format_reward": 0.3839285736903548,
+      "step": 109
+    },
+    {
+      "clip_ratio": 0.0009606969933884102,
+      "epoch": 0.016417144722727484,
+      "grad_norm": 0.04337799921631813,
+      "kl": 0.03320503234863281,
+      "learning_rate": 4.587268241878724e-06,
+      "loss": 0.0003,
+      "step": 110
+    },
+    {
+      "clip_ratio": 0.0013132303793099709,
+      "epoch": 0.0165663914929341,
+      "grad_norm": 0.03975389152765274,
+      "kl": 0.030458450317382812,
+      "learning_rate": 4.578816606352205e-06,
+      "loss": 0.0004,
+      "step": 111
+    },
+    {
+      "clip_ratio": 0.0009267619698221097,
+      "completion_length": 730.0000305175781,
+      "epoch": 0.016715638263140712,
+      "grad_norm": 0.045510776340961456,
+      "kl": 0.015173912048339844,
+      "learning_rate": 4.570288237343632e-06,
+      "loss": 0.0004,
+      "num_tokens": 6063748.0,
+      "reward": 0.5225313189439476,
+      "reward_std": 0.31967434333637357,
+      "rewards/code_reward": 0.48538846569135785,
+      "rewards/format_reward": 0.3714285744354129,
+      "step": 112
+    },
+    {
+      "clip_ratio": 0.001110001046981779,
+      "epoch": 0.016864885033347325,
+      "grad_norm": 0.04822126775979996,
+      "kl": 0.014050483703613281,
+      "learning_rate": 4.561683492686289e-06,
+      "loss": 0.0003,
+      "step": 113
+    },
+    {
+      "clip_ratio": 0.0010535481942497427,
+      "epoch": 0.01701413180355394,
+      "grad_norm": 0.047740690410137177,
+      "kl": 0.013932228088378906,
+      "learning_rate": 4.5530027334180285e-06,
+      "loss": 0.0001,
+      "step": 114
+    },
+    {
+      "clip_ratio": 0.0007933495671750279,
+      "completion_length": 842.017894744873,
+      "epoch": 0.017163378573760553,
+      "grad_norm": 0.049595024436712265,
+      "kl": 0.019130706787109375,
+      "learning_rate": 4.544246323766122e-06,
+      "loss": 0.0003,
+      "num_tokens": 6223518.0,
+      "reward": 0.40766190411522985,
+      "reward_std": 0.2016809419146739,
+      "rewards/code_reward": 0.36694761738181114,
+      "rewards/format_reward": 0.4071428570896387,
+      "step": 115
+    },
+    {
+      "clip_ratio": 0.0009076765018107835,
+      "epoch": 0.017312625343967165,
+      "grad_norm": 0.055327292531728745,
+      "kl": 0.02070760726928711,
+      "learning_rate": 4.535414631131983e-06,
+      "loss": 0.0003,
+      "step": 116
+    },
+    {
+      "clip_ratio": 0.0008779259478615131,
+      "epoch": 0.017461872114173778,
+      "grad_norm": 0.05579111725091934,
+      "kl": 0.022034168243408203,
+      "learning_rate": 4.526508026075746e-06,
+      "loss": 0.0001,
+      "step": 117
+    },
+    {
+      "clip_ratio": 0.0013046737203694647,
+      "completion_length": 891.5893230438232,
+      "epoch": 0.017611118884380394,
+      "grad_norm": 0.044059909880161285,
+      "kl": 0.02873706817626953,
+      "learning_rate": 4.517526882300721e-06,
+      "loss": 0.0004,
+      "num_tokens": 6382309.0,
+      "reward": 0.3803525387775153,
+      "reward_std": 0.21154743927763775,
+      "rewards/code_reward": 0.33945966558530927,
+      "rewards/format_reward": 0.40892857499420643,
+      "step": 118
+    },
+    {
+      "clip_ratio": 0.0011895602456206689,
+      "epoch": 0.017760365654587006,
+      "grad_norm": 0.05387920141220093,
+      "kl": 0.02957630157470703,
+      "learning_rate": 4.508471576637713e-06,
+      "loss": 0.0004,
+      "step": 119
+    },
+    {
+      "clip_ratio": 0.0011620367440627888,
+      "epoch": 0.01790961242479362,
+      "grad_norm": 0.04833860695362091,
+      "kl": 0.028537750244140625,
+      "learning_rate": 4.499342489029211e-06,
+      "loss": 0.0003,
+      "step": 120
+    },
+    {
+      "clip_ratio": 0.0010427872803120408,
+      "completion_length": 885.7411022186279,
+      "epoch": 0.018058859195000235,
+      "grad_norm": 0.04128396511077881,
+      "kl": 0.01944446563720703,
+      "learning_rate": 4.490140002513449e-06,
+      "loss": 0.0003,
+      "num_tokens": 6544827.0,
+      "reward": 0.37972693943447666,
+      "reward_std": 0.23223779880208895,
+      "rewards/code_reward": 0.3414233757648617,
+      "rewards/format_reward": 0.3830357129336335,
+      "step": 121
+    },
+    {
+      "clip_ratio": 0.0011303816172585357,
+      "epoch": 0.018208105965206847,
+      "grad_norm": 0.041573666036129,
+      "kl": 0.02045726776123047,
+      "learning_rate": 4.48086450320833e-06,
+      "loss": 0.0002,
+      "step": 122
+    },
+    {
+      "clip_ratio": 0.001024394565320108,
+      "epoch": 0.01835735273541346,
+      "grad_norm": 0.05487101525068283,
+      "kl": 0.020847320556640625,
+      "learning_rate": 4.4715163802952266e-06,
+      "loss": 0.0001,
+      "step": 123
+    },
+    {
+      "clip_ratio": 0.0011558874903130345,
+      "completion_length": 775.7053852081299,
+      "epoch": 0.018506599505620072,
+      "grad_norm": 0.04507587477564812,
+      "kl": 0.030223846435546875,
+      "learning_rate": 4.462096026002655e-06,
+      "loss": 0.0005,
+      "num_tokens": 6686845.0,
+      "reward": 0.4757464681752026,
+      "reward_std": 0.18553807557327673,
+      "rewards/code_reward": 0.433782160282135,
+      "rewards/format_reward": 0.4196428628638387,
+      "step": 124
+    },
+    {
+      "clip_ratio": 0.0010900183115154505,
+      "epoch": 0.018655846275826688,
+      "grad_norm": 0.042524147778749466,
+      "kl": 0.033313751220703125,
+      "learning_rate": 4.4526038355898144e-06,
+      "loss": 0.0003,
+      "step": 125
+    },
+    {
+      "clip_ratio": 0.0010025524697994115,
+      "epoch": 0.0188050930460333,
+      "grad_norm": 0.04530392214655876,
+      "kl": 0.03578948974609375,
+      "learning_rate": 4.4430402073300035e-06,
+      "loss": 0.0003,
+      "step": 126
+    },
+    {
+      "clip_ratio": 0.0009960982242773753,
+      "completion_length": 835.267894744873,
+      "epoch": 0.018954339816239913,
+      "grad_norm": 0.04436438903212547,
+      "kl": 0.025674819946289062,
+      "learning_rate": 4.433405542493909e-06,
+      "loss": 0.0003,
+      "num_tokens": 6840994.0,
+      "reward": 0.5514912263024598,
+      "reward_std": 0.24589035636745393,
+      "rewards/code_reward": 0.5114019273314625,
+      "rewards/format_reward": 0.40089286118745804,
+      "step": 127
+    },
+    {
+      "clip_ratio": 0.0010956393562082667,
+      "epoch": 0.01910358658644653,
+      "grad_norm": 0.046535246074199677,
+      "kl": 0.024074554443359375,
+      "learning_rate": 4.4237002453327734e-06,
+      "loss": 0.0003,
+      "step": 128
+    },
+    {
+      "clip_ratio": 0.0011050422554035322,
+      "epoch": 0.01925283335665314,
+      "grad_norm": 0.04183583706617355,
+      "kl": 0.023334503173828125,
+      "learning_rate": 4.4139247230614245e-06,
+      "loss": 0.0003,
+      "step": 129
+    },
+    {
+      "clip_ratio": 0.0010861948831006885,
+      "completion_length": 800.0625343322754,
+      "epoch": 0.019402080126859754,
+      "grad_norm": 0.04727044329047203,
+      "kl": 0.019486427307128906,
+      "learning_rate": 4.404079385841201e-06,
+      "loss": 0.0003,
+      "num_tokens": 6992950.0,
+      "reward": 0.4348937713075429,
+      "reward_std": 0.21684974804520607,
+      "rewards/code_reward": 0.3872152113035554,
+      "rewards/format_reward": 0.4767857138067484,
+      "step": 130
+    },
+    {
+      "clip_ratio": 0.0010590591746222344,
+      "epoch": 0.019551326897066366,
+      "grad_norm": 0.04723493382334709,
+      "kl": 0.019405364990234375,
+      "learning_rate": 4.394164646762734e-06,
+      "loss": 0.0001,
+      "step": 131
+    },
+    {
+      "clip_ratio": 0.0010405187804281013,
+      "epoch": 0.019700573667272982,
+      "grad_norm": 0.04620596021413803,
+      "kl": 0.020595550537109375,
+      "learning_rate": 4.384180921828618e-06,
+      "loss": 0.0001,
+      "step": 132
+    },
+    {
+      "clip_ratio": 0.0009282534629164729,
+      "completion_length": 690.0268249511719,
+      "epoch": 0.019849820437479595,
+      "grad_norm": 0.048521529883146286,
+      "kl": 0.022710800170898438,
+      "learning_rate": 4.374128629935955e-06,
+      "loss": 0.0002,
+      "num_tokens": 7130006.0,
+      "reward": 0.45080209872685373,
+      "reward_std": 0.2004464182537049,
+      "rewards/code_reward": 0.40553424460813403,
+      "rewards/format_reward": 0.45267857518047094,
+      "step": 133
+    },
+    {
+      "clip_ratio": 0.0010145993473997805,
+      "epoch": 0.019999067207686207,
+      "grad_norm": 0.04121008515357971,
+      "kl": 0.023035049438476562,
+      "learning_rate": 4.364008192858781e-06,
+      "loss": 0.0002,
+      "step": 134
+    },
+    {
+      "clip_ratio": 0.0007326995892071864,
+      "epoch": 0.020148313977892823,
+      "grad_norm": 0.04002685472369194,
+      "kl": 0.022815704345703125,
+      "learning_rate": 4.353820035230366e-06,
+      "loss": -0.0,
+      "step": 135
+    },
+    {
+      "clip_ratio": 0.0012416261888574809,
+      "completion_length": 797.4107418060303,
+      "epoch": 0.020297560748099436,
+      "grad_norm": 0.042961277067661285,
+      "kl": 0.0173187255859375,
+      "learning_rate": 4.3435645845254e-06,
+      "loss": 0.0002,
+      "num_tokens": 7274532.0,
+      "reward": 0.4839099827222526,
+      "reward_std": 0.2480696743587032,
+      "rewards/code_reward": 0.4350706939585507,
+      "rewards/format_reward": 0.48839285923168063,
+      "step": 136
+    },
+    {
+      "clip_ratio": 0.0011591103339014808,
+      "epoch": 0.020446807518306048,
+      "grad_norm": 0.042345136404037476,
+      "kl": 0.017218589782714844,
+      "learning_rate": 4.333242271042054e-06,
+      "loss": 0.0001,
+      "step": 137
+    },
+    {
+      "clip_ratio": 0.0010988163867295953,
+      "epoch": 0.020596054288512664,
+      "grad_norm": 0.03951758146286011,
+      "kl": 0.01706218719482422,
+      "learning_rate": 4.32285352788393e-06,
+      "loss": -0.0,
+      "step": 138
+    },
+    {
+      "clip_ratio": 0.0011232689575990662,
+      "completion_length": 855.330394744873,
+      "epoch": 0.020745301058719277,
+      "grad_norm": 0.0443883016705513,
+      "kl": 0.024089813232421875,
+      "learning_rate": 4.312398790941882e-06,
+      "loss": 0.0003,
+      "num_tokens": 7437319.0,
+      "reward": 0.4357127647381276,
+      "reward_std": 0.2294543270545546,
+      "rewards/code_reward": 0.3918734625913203,
+      "rewards/format_reward": 0.4383928570896387,
+      "step": 139
+    },
+    {
+      "clip_ratio": 0.001039013011904899,
+      "epoch": 0.02089454782892589,
+      "grad_norm": 0.044901248067617416,
+      "kl": 0.023738861083984375,
+      "learning_rate": 4.301878498875735e-06,
+      "loss": 0.0004,
+      "step": 140
+    },
+    {
+      "clip_ratio": 0.0010252416741423076,
+      "epoch": 0.0210437945991325,
+      "grad_norm": 0.0455520935356617,
+      "kl": 0.02365875244140625,
+      "learning_rate": 4.291293093095873e-06,
+      "loss": 0.0003,
+      "step": 141
+    },
+    {
+      "clip_ratio": 0.001033905820804648,
+      "completion_length": 854.7143211364746,
+      "epoch": 0.021193041369339118,
+      "grad_norm": 0.04243411123752594,
+      "kl": 0.03364276885986328,
+      "learning_rate": 4.280643017744723e-06,
+      "loss": 0.0004,
+      "num_tokens": 7613862.0,
+      "reward": 0.2988048434490338,
+      "reward_std": 0.20296698369202204,
+      "rewards/code_reward": 0.25853697166166967,
+      "rewards/format_reward": 0.40267857164144516,
+      "step": 142
+    },
+    {
+      "clip_ratio": 0.0012581654991663527,
+      "epoch": 0.02134228813954573,
+      "grad_norm": 0.04118216782808304,
+      "kl": 0.033367156982421875,
+      "learning_rate": 4.269928719678117e-06,
+      "loss": 0.0004,
+      "step": 143
+    },
+    {
+      "clip_ratio": 0.001068375633622054,
+      "epoch": 0.021491534909752343,
+      "grad_norm": 0.039931606501340866,
+      "kl": 0.03148937225341797,
+      "learning_rate": 4.2591506484465426e-06,
+      "loss": 0.0002,
+      "step": 144
+    },
+    {
+      "clip_ratio": 0.0013881738632335328,
+      "completion_length": 809.4286079406738,
+      "epoch": 0.02164078167995896,
+      "grad_norm": 0.05351189151406288,
+      "kl": 0.04827117919921875,
+      "learning_rate": 4.248309256276283e-06,
+      "loss": 0.0006,
+      "num_tokens": 7762821.0,
+      "reward": 0.3853453987976536,
+      "reward_std": 0.1850955591071397,
+      "rewards/code_reward": 0.3422203725203872,
+      "rewards/format_reward": 0.43125000037252903,
+      "step": 145
+    },
+    {
+      "clip_ratio": 0.0014689411145809572,
+      "epoch": 0.02179002845016557,
+      "grad_norm": 0.05481509119272232,
+      "kl": 0.04859161376953125,
+      "learning_rate": 4.23740499805044e-06,
+      "loss": 0.0006,
+      "step": 146
+    },
+    {
+      "clip_ratio": 0.0015166042630880838,
+      "epoch": 0.021939275220372183,
+      "grad_norm": 0.047822095453739166,
+      "kl": 0.039295196533203125,
+      "learning_rate": 4.22643833128985e-06,
+      "loss": 0.0005,
+      "step": 147
+    },
+    {
+      "clip_ratio": 0.0014342169270094018,
+      "completion_length": 730.4107513427734,
+      "epoch": 0.022088521990578796,
+      "grad_norm": 0.04696185141801834,
+      "kl": 0.02483367919921875,
+      "learning_rate": 4.215409716133885e-06,
+      "loss": 0.0003,
+      "num_tokens": 7900536.0,
+      "reward": 0.48524553480092436,
+      "reward_std": 0.20289410289842635,
+      "rewards/code_reward": 0.43283481095568277,
+      "rewards/format_reward": 0.5241071395576,
+      "step": 148
+    },
+    {
+      "clip_ratio": 0.0014147953188512474,
+      "epoch": 0.022237768760785412,
+      "grad_norm": 0.049281951040029526,
+      "kl": 0.025539398193359375,
+      "learning_rate": 4.204319615321151e-06,
+      "loss": 0.0002,
+      "step": 149
+    },
+    {
+      "clip_ratio": 0.001617625761355157,
+      "epoch": 0.022387015530992024,
+      "grad_norm": 0.047861773520708084,
+      "kl": 0.026336669921875,
+      "learning_rate": 4.193168494170065e-06,
+      "loss": 0.0003,
+      "step": 150
+    },
+    {
+      "clip_ratio": 0.0011637477964541176,
+      "completion_length": 847.5804023742676,
+      "epoch": 0.022536262301198637,
+      "grad_norm": 0.043306585401296616,
+      "kl": 0.01751708984375,
+      "learning_rate": 4.181956820559339e-06,
+      "loss": 0.0002,
+      "num_tokens": 8062014.0,
+      "reward": 0.4883292701561004,
+      "reward_std": 0.24242727342061698,
+      "rewards/code_reward": 0.4366328134201467,
+      "rewards/format_reward": 0.5169642921537161,
+      "step": 151
+    },
+    {
+      "clip_ratio": 0.0012195148992759641,
+      "epoch": 0.022685509071405253,
+      "grad_norm": 0.04269765689969063,
+      "kl": 0.018365859985351562,
+      "learning_rate": 4.170685064908342e-06,
+      "loss": 0.0001,
+      "step": 152
+    },
+    {
+      "clip_ratio": 0.0012396183246892178,
+      "epoch": 0.022834755841611865,
+      "grad_norm": 0.061775047332048416,
+      "kl": 0.018674850463867188,
+      "learning_rate": 4.159353700157365e-06,
+      "loss": 0.0001,
+      "step": 153
+    },
+    {
+      "clip_ratio": 0.0012902800481242593,
+      "completion_length": 876.3750400543213,
+      "epoch": 0.022984002611818478,
+      "grad_norm": 0.04410018399357796,
+      "kl": 0.02848052978515625,
+      "learning_rate": 4.14796320174778e-06,
+      "loss": 0.0004,
+      "num_tokens": 8221495.0,
+      "reward": 0.5067510595545173,
+      "reward_std": 0.1769608927424997,
+      "rewards/code_reward": 0.45487602904904634,
+      "rewards/format_reward": 0.5187500035390258,
+      "step": 154
+    },
+    {
+      "clip_ratio": 0.0012712476654996863,
+      "epoch": 0.023133249382025094,
+      "grad_norm": 0.042361412197351456,
+      "kl": 0.02883148193359375,
+      "learning_rate": 4.136514047602087e-06,
+      "loss": 0.0003,
+      "step": 155
+    },
+    {
+      "clip_ratio": 0.0012503136294981232,
+      "epoch": 0.023282496152231706,
+      "grad_norm": 0.04951699078083038,
+      "kl": 0.030870437622070312,
+      "learning_rate": 4.1250067181038635e-06,
+      "loss": 0.0003,
+      "step": 156
+    },
+    {
+      "clip_ratio": 0.001006045729809557,
+      "completion_length": 770.8571758270264,
+      "epoch": 0.02343174292243832,
+      "grad_norm": 0.19089701771736145,
+      "kl": 0.158905029296875,
+      "learning_rate": 4.113441696077608e-06,
+      "loss": 0.0015,
+      "num_tokens": 8361654.0,
+      "reward": 0.5336465365253389,
+      "reward_std": 0.22312293585855514,
+      "rewards/code_reward": 0.4696286738035269,
+      "rewards/format_reward": 0.6401785779744387,
+      "step": 157
+    },
+    {
+      "clip_ratio": 0.000968805668890127,
+      "epoch": 0.02358098969264493,
+      "grad_norm": 0.07467727363109589,
+      "kl": 0.08309364318847656,
+      "learning_rate": 4.101819466768484e-06,
+      "loss": 0.0009,
+      "step": 158
+    },
+    {
+      "clip_ratio": 0.001212779263369157,
+      "epoch": 0.023730236462851547,
+      "grad_norm": 0.10417937487363815,
+      "kl": 0.049536705017089844,
+      "learning_rate": 4.0901405178219535e-06,
+      "loss": 0.0007,
+      "step": 159
+    },
+    {
+      "clip_ratio": 0.0012088071471225703,
+      "completion_length": 773.2321720123291,
+      "epoch": 0.02387948323305816,
+      "grad_norm": 0.0834612250328064,
+      "kl": 0.07547187805175781,
+      "learning_rate": 4.078405339263326e-06,
+      "loss": 0.0008,
+      "num_tokens": 8509170.0,
+      "reward": 0.5869864404667169,
+      "reward_std": 0.19239688021480106,
+      "rewards/code_reward": 0.5358257133048028,
+      "rewards/format_reward": 0.511607144959271,
+      "step": 160
+    },
+    {
+      "clip_ratio": 0.001222252394654788,
+      "epoch": 0.024028730003264772,
+      "grad_norm": 0.11993829160928726,
+      "kl": 0.057216644287109375,
+      "learning_rate": 4.06661442347719e-06,
+      "loss": 0.0005,
+      "step": 161
+    },
+    {
+      "clip_ratio": 0.001276199891435681,
+      "epoch": 0.024177976773471388,
+      "grad_norm": 0.05552421510219574,
+      "kl": 0.04888153076171875,
+      "learning_rate": 4.054768265186758e-06,
+      "loss": 0.0005,
+      "step": 162
+    },
+    {
+      "clip_ratio": 0.0010032035952463048,
+      "completion_length": 790.5089588165283,
+      "epoch": 0.024327223543678,
+      "grad_norm": 0.1210794448852539,
+      "kl": 0.03442573547363281,
+      "learning_rate": 4.0428673614331036e-06,
+      "loss": 0.0003,
+      "num_tokens": 8655443.0,
+      "reward": 0.5846006067004055,
+      "reward_std": 0.3078883128473535,
+      "rewards/code_reward": 0.5278148755605798,
+      "rewards/format_reward": 0.5678571437019855,
+      "step": 163
+    },
+    {
+      "clip_ratio": 0.0010885789324674988,
+      "epoch": 0.024476470313884613,
+      "grad_norm": 0.04959575831890106,
+      "kl": 0.032726287841796875,
+      "learning_rate": 4.030912211554316e-06,
+      "loss": 0.0003,
+      "step": 164
+    },
+    {
+      "clip_ratio": 0.0011579382626223378,
+      "epoch": 0.024625717084091225,
+      "grad_norm": 0.0483749583363533,
+      "kl": 0.030529022216796875,
+      "learning_rate": 4.018903317164539e-06,
+      "loss": 0.0003,
+      "step": 165
+    },
+    {
+      "clip_ratio": 0.001137841476520407,
+      "completion_length": 954.4554138183594,
+      "epoch": 0.02477496385429784,
+      "grad_norm": 0.04242149367928505,
+      "kl": 0.02735614776611328,
+      "learning_rate": 4.006841182132932e-06,
+      "loss": 0.0003,
+      "num_tokens": 8825811.0,
+      "reward": 0.343663371168077,
+      "reward_std": 0.21622509372537024,
+      "rewards/code_reward": 0.2994669284671545,
+      "rewards/format_reward": 0.4419642901048064,
+      "step": 166
+    },
+    {
+      "clip_ratio": 0.001333304300715099,
+      "epoch": 0.024924210624504454,
+      "grad_norm": 0.05318411439657211,
+      "kl": 0.02469921112060547,
+      "learning_rate": 3.9947263125625195e-06,
+      "loss": 0.0004,
+      "step": 167
+    },
+    {
+      "clip_ratio": 0.0014823580131633207,
+      "epoch": 0.025073457394711066,
+      "grad_norm": 0.04331374540925026,
+      "kl": 0.023836135864257812,
+      "learning_rate": 3.982559216768967e-06,
+      "loss": 0.0003,
+      "step": 168
+    },
+    {
+      "clip_ratio": 0.001230707457580138,
+      "completion_length": 1086.4554080963135,
+      "epoch": 0.025222704164917682,
+      "grad_norm": 0.043598975986242294,
+      "kl": 0.014312744140625,
+      "learning_rate": 3.970340405259245e-06,
+      "loss": 0.0003,
+      "num_tokens": 9021743.0,
+      "reward": 0.2815680430503562,
+      "reward_std": 0.18689183273818344,
+      "rewards/code_reward": 0.22210374718997627,
+      "rewards/format_reward": 0.594642854295671,
+      "step": 169
+    },
+    {
+      "clip_ratio": 0.0012951591161254328,
+      "epoch": 0.025371950935124295,
+      "grad_norm": 0.045120351016521454,
+      "kl": 0.014659881591796875,
+      "learning_rate": 3.958070390710214e-06,
+      "loss": 0.0001,
+      "step": 170
+    },
+    {
+      "clip_ratio": 0.0013625593310280237,
+      "epoch": 0.025521197705330907,
+      "grad_norm": 0.042519036680459976,
+      "kl": 0.014459609985351562,
+      "learning_rate": 3.945749687947109e-06,
+      "loss": 0.0002,
+      "step": 171
+    },
+    {
+      "clip_ratio": 0.001266109573407448,
+      "completion_length": 758.3928909301758,
+      "epoch": 0.025670444475537523,
+      "grad_norm": 0.04733174666762352,
+      "kl": 0.027559280395507812,
+      "learning_rate": 3.933378813921942e-06,
+      "loss": 0.0004,
+      "num_tokens": 9178762.0,
+      "reward": 0.5201502540148795,
+      "reward_std": 0.27701563900336623,
+      "rewards/code_reward": 0.4503288194537163,
+      "rewards/format_reward": 0.6982142943888903,
+      "step": 172
+    },
+    {
+      "clip_ratio": 0.0013183873506932287,
+      "epoch": 0.025819691245744136,
+      "grad_norm": 0.043405089527368546,
+      "kl": 0.028142929077148438,
+      "learning_rate": 3.920958287691811e-06,
+      "loss": 0.0003,
+      "step": 173
+    },
+    {
+      "clip_ratio": 0.0013744723346462706,
+      "epoch": 0.025968938015950748,
+      "grad_norm": 0.04960908368229866,
+      "kl": 0.030193328857421875,
+      "learning_rate": 3.908488630397121e-06,
+      "loss": 0.0003,
+      "step": 174
+    },
+    {
+      "clip_ratio": 0.0011061989516747417,
+      "completion_length": 970.54469871521,
+      "epoch": 0.02611818478615736,
+      "grad_norm": 0.04485900327563286,
+      "kl": 0.01617908477783203,
+      "learning_rate": 3.8959703652397175e-06,
+      "loss": 0.0001,
+      "num_tokens": 9359647.0,
+      "reward": 0.34467554092407227,
+      "reward_std": 0.18839130480773747,
+      "rewards/code_reward": 0.2806576928123832,
+      "rewards/format_reward": 0.6401785714551806,
+      "step": 175
+    },
+    {
+      "clip_ratio": 0.001184371045383159,
+      "epoch": 0.026267431556363977,
+      "grad_norm": 0.05172208696603775,
+      "kl": 0.01747894287109375,
+      "learning_rate": 3.883404017460935e-06,
+      "loss": 0.0001,
+      "step": 176
+    },
+    {
+      "clip_ratio": 0.0011093282719230046,
+      "epoch": 0.02641667832657059,
+      "grad_norm": 0.0427677184343338,
+      "kl": 0.019744873046875,
+      "learning_rate": 3.870790114319559e-06,
+      "loss": 0.0,
+      "step": 177
+    },
+    {
+      "clip_ratio": 0.001037665282638045,
+      "completion_length": 806.4732513427734,
+      "epoch": 0.0265659250967772,
+      "grad_norm": 0.04383108392357826,
+      "kl": 0.03477668762207031,
+      "learning_rate": 3.858129185069701e-06,
+      "loss": 0.0004,
+      "num_tokens": 9501968.0,
+      "reward": 0.5641923944931477,
+      "reward_std": 0.20727904001250863,
+      "rewards/code_reward": 0.4969602590499562,
+      "rewards/format_reward": 0.6723214294761419,
+      "step": 178
+    },
+    {
+      "clip_ratio": 0.0013952661065559369,
+      "epoch": 0.026715171866983817,
+      "grad_norm": 0.04694554954767227,
+      "kl": 0.036670684814453125,
+      "learning_rate": 3.845421760938597e-06,
+      "loss": 0.0004,
+      "step": 179
+    },
+    {
+      "clip_ratio": 0.0012856538778578397,
+      "epoch": 0.02686441863719043,
+      "grad_norm": 0.04348964989185333,
+      "kl": 0.03646087646484375,
+      "learning_rate": 3.832668375104312e-06,
+      "loss": 0.0003,
+      "step": 180
+    },
+    {
+      "clip_ratio": 0.0012388697778078495,
+      "completion_length": 785.8928909301758,
+      "epoch": 0.027013665407397042,
+      "grad_norm": 0.04777670279145241,
+      "kl": 0.034305572509765625,
+      "learning_rate": 3.8198695626733725e-06,
+      "loss": 0.0004,
+      "num_tokens": 9653849.0,
+      "reward": 0.584011502796784,
+      "reward_std": 0.22272017179057002,
+      "rewards/code_reward": 0.49740435276180506,
+      "rewards/format_reward": 0.8660714142024517,
+      "step": 181
+    },
+    {
+      "clip_ratio": 0.0014184717001626268,
+      "epoch": 0.027162912177603655,
+      "grad_norm": 0.06207079812884331,
+      "kl": 0.032253265380859375,
+      "learning_rate": 3.8070258606583156e-06,
+      "loss": 0.0003,
+      "step": 182
+    },
+    {
+      "clip_ratio": 0.0012104590714443475,
+      "epoch": 0.02731215894781027,
+      "grad_norm": 0.047974348068237305,
+      "kl": 0.03397369384765625,
+      "learning_rate": 3.7941378079551544e-06,
+      "loss": 0.0003,
+      "step": 183
+    },
+    {
+      "clip_ratio": 0.0013559910148615018,
+      "completion_length": 798.0179023742676,
+      "epoch": 0.027461405718016883,
+      "grad_norm": 0.0428900420665741,
+      "kl": 0.028224945068359375,
+      "learning_rate": 3.7812059453207677e-06,
+      "loss": 0.0004,
+      "num_tokens": 9802783.0,
+      "reward": 0.5101535907015204,
+      "reward_std": 0.2436997564509511,
+      "rewards/code_reward": 0.4328321506618522,
+      "rewards/format_reward": 0.7732142838649452,
+      "step": 184
+    },
+    {
+      "clip_ratio": 0.0014902301008987706,
+      "epoch": 0.027610652488223496,
+      "grad_norm": 0.04620678350329399,
+      "kl": 0.0277099609375,
+      "learning_rate": 3.768230815350213e-06,
+      "loss": 0.0003,
+      "step": 185
+    },
+    {
+      "clip_ratio": 0.0015550048210570822,
+      "epoch": 0.02775989925843011,
+      "grad_norm": 0.04612104594707489,
+      "kl": 0.029872894287109375,
+      "learning_rate": 3.7552129624539557e-06,
+      "loss": 0.0003,
+      "step": 186
+    },
+    {
+      "clip_ratio": 0.00117877590128046,
+      "completion_length": 909.267894744873,
+      "epoch": 0.027909146028636724,
+      "grad_norm": 0.047490544617176056,
+      "kl": 0.035709381103515625,
+      "learning_rate": 3.7421529328350316e-06,
+      "loss": 0.0004,
+      "num_tokens": 9970498.0,
+      "reward": 0.3525247387588024,
+      "reward_std": 0.14782223524525762,
+      "rewards/code_reward": 0.2900247216457501,
+      "rewards/format_reward": 0.6250000037252903,
+      "step": 187
+    },
+    {
+      "clip_ratio": 0.0012039170615025796,
+      "epoch": 0.028058392798843337,
+      "grad_norm": 0.04097956046462059,
+      "kl": 0.037628173828125,
+      "learning_rate": 3.7290512744661274e-06,
+      "loss": 0.0004,
+      "step": 188
+    },
+    {
+      "clip_ratio": 0.00120592799885344,
+      "epoch": 0.028207639569049953,
+      "grad_norm": 0.04101557284593582,
+      "kl": 0.03668975830078125,
+      "learning_rate": 3.715908537066589e-06,
+      "loss": 0.0003,
+      "step": 189
+    },
+    {
+      "clip_ratio": 0.0012375264723232249,
+      "completion_length": 972.2857627868652,
+      "epoch": 0.028356886339256565,
+      "grad_norm": 0.0460306815803051,
+      "kl": 0.035683631896972656,
+      "learning_rate": 3.7027252720793538e-06,
+      "loss": 0.0006,
+      "num_tokens": 10154728.0,
+      "reward": 0.4762301933951676,
+      "reward_std": 0.2230017096735537,
+      "rewards/code_reward": 0.4048016065498814,
+      "rewards/format_reward": 0.7142857164144516,
+      "step": 190
+    },
+    {
+      "clip_ratio": 0.0012021254078717902,
+      "epoch": 0.028506133109463178,
+      "grad_norm": 0.053284067660570145,
+      "kl": 0.034501075744628906,
+      "learning_rate": 3.689502032647817e-06,
+      "loss": 0.0005,
+      "step": 191
+    },
+    {
+      "clip_ratio": 0.0013434358115773648,
+      "epoch": 0.02865537987966979,
+      "grad_norm": 0.044003624469041824,
+      "kl": 0.031785011291503906,
+      "learning_rate": 3.6762393735926245e-06,
+      "loss": 0.0004,
+      "step": 192
+    },
+    {
+      "clip_ratio": 0.001241399731952697,
+      "completion_length": 853.8482551574707,
+      "epoch": 0.028804626649876406,
+      "grad_norm": 0.04587972164154053,
+      "kl": 0.0348663330078125,
+      "learning_rate": 3.6629378513883852e-06,
+      "loss": 0.0004,
+      "num_tokens": 10319365.0,
+      "reward": 0.4278743101749569,
+      "reward_std": 0.16044311877340078,
+      "rewards/code_reward": 0.3440350177406799,
+      "rewards/format_reward": 0.8383928369730711,
+      "step": 193
+    },
+    {
+      "clip_ratio": 0.0014909484089002945,
+      "epoch": 0.02895387342008302,
+      "grad_norm": 0.04408523812890053,
+      "kl": 0.03650093078613281,
+      "learning_rate": 3.6495980241403307e-06,
+      "loss": 0.0004,
+      "step": 194
+    },
+    {
+      "clip_ratio": 0.0011587896278797416,
+      "epoch": 0.02910312019028963,
+      "grad_norm": 0.044679079204797745,
+      "kl": 0.037075042724609375,
+      "learning_rate": 3.636220451560896e-06,
+      "loss": 0.0003,
+      "step": 195
+    },
+    {
+      "clip_ratio": 0.0008984732257886208,
+      "completion_length": 807.0536079406738,
+      "epoch": 0.029252366960496247,
+      "grad_norm": 0.04803880304098129,
+      "kl": 0.03685188293457031,
+      "learning_rate": 3.622805694946235e-06,
+      "loss": 0.0004,
+      "num_tokens": 10472860.0,
+      "reward": 0.4805865343660116,
+      "reward_std": 0.2713788175024092,
+      "rewards/code_reward": 0.4047829551855102,
+      "rewards/format_reward": 0.7580357203260064,
+      "step": 196
+    },
+    {
+      "clip_ratio": 0.000974786184997356,
+      "epoch": 0.02940161373070286,
+      "grad_norm": 0.0561777763068676,
+      "kl": 0.0364837646484375,
+      "learning_rate": 3.609354317152667e-06,
+      "loss": 0.0003,
+      "step": 197
+    },
+    {
+      "clip_ratio": 0.0010943580300590838,
+      "epoch": 0.029550860500909472,
+      "grad_norm": 0.04243557155132294,
+      "kl": 0.03566169738769531,
+      "learning_rate": 3.595866882573063e-06,
+      "loss": 0.0003,
+      "step": 198
+    },
+    {
+      "clip_ratio": 0.001358222019916866,
+      "completion_length": 625.1071681976318,
+      "epoch": 0.029700107271116084,
+      "grad_norm": 0.060445670038461685,
+      "kl": 0.04364013671875,
+      "learning_rate": 3.5823439571131675e-06,
+      "loss": 0.0004,
+      "num_tokens": 10594252.0,
+      "reward": 0.5816141716204584,
+      "reward_std": 0.24522001342847943,
+      "rewards/code_reward": 0.4959891536273062,
+      "rewards/format_reward": 0.8562499936670065,
+      "step": 199
+    },
+    {
+      "clip_ratio": 0.0011922506128030363,
+      "epoch": 0.0298493540413227,
+      "grad_norm": 0.09876689314842224,
+      "kl": 0.04461669921875,
+      "learning_rate": 3.5687861081678477e-06,
+      "loss": 0.0004,
+      "step": 200
+    },
+    {
+      "clip_ratio": 0.0014354662271216512,
+      "epoch": 0.029998600811529313,
+      "grad_norm": 0.0558474026620388,
+      "kl": 0.0430145263671875,
+      "learning_rate": 3.555193904597291e-06,
+      "loss": 0.0004,
+      "step": 201
+    },
+    {
+      "clip_ratio": 0.0016133483604789944,
+      "completion_length": 778.0803890228271,
+      "epoch": 0.030147847581735925,
+      "grad_norm": 0.0491982102394104,
+      "kl": 0.038715362548828125,
+      "learning_rate": 3.541567916703138e-06,
+      "loss": 0.0006,
+      "num_tokens": 10747848.0,
+      "reward": 0.4137029910925776,
+      "reward_std": 0.23209618078544736,
+      "rewards/code_reward": 0.33709583431482315,
+      "rewards/format_reward": 0.7660714257508516,
+      "step": 202
+    },
+    {
+      "clip_ratio": 0.0014115704216237646,
+      "epoch": 0.03029709435194254,
+      "grad_norm": 0.04640298709273338,
+      "kl": 0.039661407470703125,
+      "learning_rate": 3.5279087162045517e-06,
+      "loss": 0.0005,
+      "step": 203
+    },
+    {
+      "clip_ratio": 0.001271567817639152,
+      "epoch": 0.030446341122149154,
+      "grad_norm": 0.044124212116003036,
+      "kl": 0.042690277099609375,
+      "learning_rate": 3.5142168762142265e-06,
+      "loss": 0.0003,
+      "step": 204
+    },
+    {
+      "clip_ratio": 0.0009439818841201486,
+      "completion_length": 891.4286041259766,
+      "epoch": 0.030595587892355766,
+      "grad_norm": 0.04894348606467247,
+      "kl": 0.032591819763183594,
+      "learning_rate": 3.500492971214347e-06,
+      "loss": 0.0004,
+      "num_tokens": 10915987.0,
+      "reward": 0.5286302748136222,
+      "reward_std": 0.26153607084415853,
+      "rewards/code_reward": 0.44514812077977695,
+      "rewards/format_reward": 0.8348214142024517,
+      "step": 205
+    },
+    {
+      "clip_ratio": 0.0010812356740643736,
+      "epoch": 0.030744834662562382,
+      "grad_norm": 0.04213985055685043,
+      "kl": 0.033928871154785156,
+      "learning_rate": 3.48673757703248e-06,
+      "loss": 0.0004,
+      "step": 206
+    },
+    {
+      "clip_ratio": 0.0011472811147541506,
+      "epoch": 0.030894081432768995,
+      "grad_norm": 0.05198606848716736,
+      "kl": 0.03620433807373047,
+      "learning_rate": 3.472951270817418e-06,
+      "loss": 0.0003,
+      "step": 207
+    },
+    {
+      "clip_ratio": 0.001109095293941209,
+      "completion_length": 705.3571720123291,
+      "epoch": 0.031043328202975607,
+      "grad_norm": 0.04343515262007713,
+      "kl": 0.0405426025390625,
+      "learning_rate": 3.4591346310149578e-06,
+      "loss": 0.0004,
+      "num_tokens": 11062166.0,
+      "reward": 0.634864809922874,
+      "reward_std": 0.21623162645846605,
+      "rewards/code_reward": 0.5361148179508746,
+      "rewards/format_reward": 0.9874999988824129,
+      "step": 208
+    },
+    {
+      "clip_ratio": 0.0008893829726730473,
+      "epoch": 0.03119257497318222,
+      "grad_norm": 0.04344526305794716,
+      "kl": 0.039653778076171875,
+      "learning_rate": 3.445288237343632e-06,
+      "loss": 0.0003,
+      "step": 209
+    },
+    {
+      "clip_ratio": 0.0009056666758624488,
+      "epoch": 0.031341821743388835,
+      "grad_norm": 0.04229241609573364,
+      "kl": 0.038776397705078125,
+      "learning_rate": 3.4314126707703895e-06,
+      "loss": 0.0002,
+      "step": 210
+    },
+    {
+      "clip_ratio": 0.0015824516594875604,
+      "completion_length": 766.2411098480225,
+      "epoch": 0.03149106851359545,
+      "grad_norm": 0.04466083645820618,
+      "kl": 0.044189453125,
+      "learning_rate": 3.4175085134862128e-06,
+      "loss": 0.0006,
+      "num_tokens": 11210110.0,
+      "reward": 0.6619070116430521,
+      "reward_std": 0.2607162531930953,
+      "rewards/code_reward": 0.5710141453891993,
+      "rewards/format_reward": 0.9089285656809807,
+      "step": 211
+    },
+    {
+      "clip_ratio": 0.0015846481364860665,
+      "epoch": 0.03164031528380206,
+      "grad_norm": 0.044882841408252716,
+      "kl": 0.0458831787109375,
+      "learning_rate": 3.4035763488816953e-06,
+      "loss": 0.0006,
+      "step": 212
+    },
+    {
+      "clip_ratio": 0.0014327153403428383,
+      "epoch": 0.03178956205400867,
+      "grad_norm": 0.0454876646399498,
+      "kl": 0.04427337646484375,
+      "learning_rate": 3.3896167615225594e-06,
+      "loss": 0.0005,
+      "step": 213
+    },
+    {
+      "clip_ratio": 0.0010587101842247648,
+      "completion_length": 918.2232494354248,
+      "epoch": 0.031938808824215285,
+      "grad_norm": 0.05162033438682556,
+      "kl": 0.045897483825683594,
+      "learning_rate": 3.375630337125133e-06,
+      "loss": 0.0004,
+      "num_tokens": 11378128.0,
+      "reward": 0.5001084273681045,
+      "reward_std": 0.2754337468650192,
+      "rewards/code_reward": 0.4088584103010362,
+      "rewards/format_reward": 0.9124999912455678,
+      "step": 214
+    },
+    {
+      "clip_ratio": 0.0010744722167146392,
+      "epoch": 0.032088055594421905,
+      "grad_norm": 0.04861937463283539,
+      "kl": 0.042781829833984375,
+      "learning_rate": 3.361617662531772e-06,
+      "loss": 0.0003,
+      "step": 215
+    },
+    {
+      "clip_ratio": 0.0014411782067327294,
+      "epoch": 0.03223730236462852,
+      "grad_norm": 0.04841092228889465,
+      "kl": 0.04188394546508789,
+      "learning_rate": 3.347579325686237e-06,
+      "loss": 0.0002,
+      "step": 216
+    },
+    {
+      "clip_ratio": 0.001376153410092229,
+      "completion_length": 905.0268306732178,
+      "epoch": 0.03238654913483513,
+      "grad_norm": 0.04793982580304146,
+      "kl": 0.03996849060058594,
+      "learning_rate": 3.333515915609027e-06,
+      "loss": 0.0006,
+      "num_tokens": 11546320.0,
+      "reward": 0.3640421375166625,
+      "reward_std": 0.0876690277364105,
+      "rewards/code_reward": 0.27386354363989085,
+      "rewards/format_reward": 0.901785722002387,
+      "step": 217
+    },
+    {
+      "clip_ratio": 0.001170761366665829,
+      "epoch": 0.03253579590504174,
+      "grad_norm": 0.04501071572303772,
+      "kl": 0.04147529602050781,
+      "learning_rate": 3.3194280223726616e-06,
+      "loss": 0.0006,
+      "step": 218
+    },
+    {
+      "clip_ratio": 0.0011851877379740472,
+      "epoch": 0.032685042675248355,
+      "grad_norm": 0.04706224054098129,
+      "kl": 0.041217803955078125,
+      "learning_rate": 3.305316237076927e-06,
+      "loss": 0.0004,
+      "step": 219
+    },
+    {
+      "clip_ratio": 0.0012976112229807768,
+      "completion_length": 814.6964626312256,
+      "epoch": 0.03283428944545497,
+      "grad_norm": 0.041682858020067215,
+      "kl": 0.037784576416015625,
+      "learning_rate": 3.291181151824071e-06,
+      "loss": 0.0005,
+      "num_tokens": 11696695.0,
+      "reward": 0.5471810144372284,
+      "reward_std": 0.14302252768538892,
+      "rewards/code_reward": 0.45718098066572566,
+      "rewards/format_reward": 0.8999999836087227,
+      "step": 220
+    },
+    {
+      "clip_ratio": 0.0012087585309927817,
+      "epoch": 0.03298353621566158,
+      "grad_norm": 0.042474351823329926,
+      "kl": 0.04036712646484375,
+      "learning_rate": 3.27702335969396e-06,
+      "loss": 0.0005,
+      "step": 221
+    },
+    {
+      "clip_ratio": 0.0015525895069004036,
+      "epoch": 0.0331327829858682,
+      "grad_norm": 0.041807934641838074,
+      "kl": 0.040065765380859375,
+      "learning_rate": 3.2628434547191985e-06,
+      "loss": 0.0005,
+      "step": 222
+    },
+    {
+      "clip_ratio": 0.0013409430903266184,
+      "completion_length": 868.7321815490723,
+      "epoch": 0.03328202975607481,
+      "grad_norm": 0.047505587339401245,
+      "kl": 0.036075592041015625,
+      "learning_rate": 3.2486420318601973e-06,
+      "loss": 0.0006,
+      "num_tokens": 11861025.0,
+      "reward": 0.3259413610212505,
+      "reward_std": 0.23739346489310265,
+      "rewards/code_reward": 0.23451278542052023,
+      "rewards/format_reward": 0.9142857156693935,
+      "step": 223
+    },
+    {
+      "clip_ratio": 0.00125404930440709,
+      "epoch": 0.033431276526281424,
+      "grad_norm": 0.05282680317759514,
+      "kl": 0.03722381591796875,
+      "learning_rate": 3.2344196869802187e-06,
+      "loss": 0.0005,
+      "step": 224
+    },
+    {
+      "clip_ratio": 0.0013416263791441452,
+      "epoch": 0.033580523296488037,
+      "grad_norm": 0.0429045669734478,
+      "kl": 0.04016876220703125,
+      "learning_rate": 3.2201770168203694e-06,
+      "loss": 0.0005,
+      "step": 225
+    },
+    {
+      "clip_ratio": 0.0010339052696508588,
+      "completion_length": 867.5000495910645,
+      "epoch": 0.03372977006669465,
+      "grad_norm": 0.04467302933335304,
+      "kl": 0.04944610595703125,
+      "learning_rate": 3.205914618974563e-06,
+      "loss": 0.0004,
+      "num_tokens": 12023194.0,
+      "reward": 0.30351110571064055,
+      "reward_std": 0.16383928316645324,
+      "rewards/code_reward": 0.20235038634564262,
+      "rewards/format_reward": 1.0116071458905935,
+      "step": 226
+    },
+    {
+      "clip_ratio": 0.0010862592589546693,
+      "epoch": 0.03387901683690126,
+      "grad_norm": 0.043438930064439774,
+      "kl": 0.04763984680175781,
+      "learning_rate": 3.1916330918644496e-06,
+      "loss": 0.0005,
+      "step": 227
+    },
+    {
+      "clip_ratio": 0.0011292613503428583,
+      "epoch": 0.03402826360710788,
+      "grad_norm": 0.04707575961947441,
+      "kl": 0.047130584716796875,
+      "learning_rate": 3.177333034714303e-06,
+      "loss": 0.0004,
+      "step": 228
+    },
+    {
+      "clip_ratio": 0.0009916176568367518,
+      "completion_length": 861.1518287658691,
+      "epoch": 0.03417751037731449,
+      "grad_norm": 0.11236114799976349,
+      "kl": 0.043548583984375,
+      "learning_rate": 3.1630150475258813e-06,
+      "loss": 0.0004,
+      "num_tokens": 12187333.0,
+      "reward": 0.6236218712292612,
+      "reward_std": 0.2065372944343835,
+      "rewards/code_reward": 0.5138004226610065,
+      "rewards/format_reward": 1.0982142873108387,
+      "step": 229
+    },
+    {
+      "clip_ratio": 0.0009899907190629165,
+      "epoch": 0.034326757147521106,
+      "grad_norm": 0.07119898498058319,
+      "kl": 0.043308258056640625,
+      "learning_rate": 3.148679731053252e-06,
+      "loss": 0.0005,
+      "step": 230
+    },
+    {
+      "clip_ratio": 0.0009169172490146593,
+      "epoch": 0.03447600391772772,
+      "grad_norm": 0.045384377241134644,
+      "kl": 0.044551849365234375,
+      "learning_rate": 3.1343276867775805e-06,
+      "loss": 0.0003,
+      "step": 231
+    },
+    {
+      "clip_ratio": 0.001247574469743995,
+      "completion_length": 836.8482475280762,
+      "epoch": 0.03462525068793433,
+      "grad_norm": 0.045782655477523804,
+      "kl": 0.051326751708984375,
+      "learning_rate": 3.1199595168819043e-06,
+      "loss": 0.0005,
+      "num_tokens": 12342541.0,
+      "reward": 0.5260538989678025,
+      "reward_std": 0.2387256936635822,
+      "rewards/code_reward": 0.4242681765317684,
+      "rewards/format_reward": 1.0178571455180645,
+      "step": 232
+    },
+    {
+      "clip_ratio": 0.0010094128338096198,
+      "epoch": 0.03477449745814094,
+      "grad_norm": 0.04860462620854378,
+      "kl": 0.051258087158203125,
+      "learning_rate": 3.105575824225852e-06,
+      "loss": 0.0004,
+      "step": 233
+    },
+    {
+      "clip_ratio": 0.001277090563235106,
+      "epoch": 0.034923744228347556,
+      "grad_norm": 0.04626248776912689,
+      "kl": 0.052829742431640625,
+      "learning_rate": 3.091177212320363e-06,
+      "loss": 0.0004,
+      "step": 234
+    },
+    {
+      "clip_ratio": 0.0012918142674607225,
+      "completion_length": 863.321455001831,
+      "epoch": 0.035072990998554175,
+      "grad_norm": 0.060307588428258896,
+      "kl": 0.0425262451171875,
+      "learning_rate": 3.0767642853023538e-06,
+      "loss": 0.0004,
+      "num_tokens": 12496871.0,
+      "reward": 0.5240239482372999,
+      "reward_std": 0.1900554287713021,
+      "rewards/code_reward": 0.4305418040603399,
+      "rewards/format_reward": 0.9348214156925678,
+      "step": 235
+    },
+    {
+      "clip_ratio": 0.0013760660076513886,
+      "epoch": 0.03522223776876079,
+      "grad_norm": 0.04616798833012581,
+      "kl": 0.040264129638671875,
+      "learning_rate": 3.062337647909376e-06,
+      "loss": 0.0003,
+      "step": 236
+    },
+    {
+      "clip_ratio": 0.001137408380600391,
+      "epoch": 0.0353714845389674,
+      "grad_norm": 0.04546603932976723,
+      "kl": 0.040740966796875,
+      "learning_rate": 3.04789790545424e-06,
+      "loss": 0.0003,
+      "step": 237
+    },
+    {
+      "clip_ratio": 0.001182597396109486,
+      "completion_length": 875.6161231994629,
+      "epoch": 0.03552073130917401,
+      "grad_norm": 0.04508831351995468,
+      "kl": 0.0308074951171875,
+      "learning_rate": 3.033445663799621e-06,
+      "loss": 0.0003,
+      "num_tokens": 12661181.0,
+      "reward": 0.5413230061531067,
+      "reward_std": 0.23713362775743008,
+      "rewards/code_reward": 0.44221584234037437,
+      "rewards/format_reward": 0.9910714235156775,
+      "step": 238
+    },
+    {
+      "clip_ratio": 0.001146262224210659,
+      "epoch": 0.035669978079380625,
+      "grad_norm": 0.04483221843838692,
+      "kl": 0.031106948852539062,
+      "learning_rate": 3.018981529332633e-06,
+      "loss": 0.0004,
+      "step": 239
+    },
+    {
+      "clip_ratio": 0.0010552481926424662,
+      "epoch": 0.03581922484958724,
+      "grad_norm": 0.04423677921295166,
+      "kl": 0.030683517456054688,
+      "learning_rate": 3.00450610893939e-06,
+      "loss": 0.0003,
+      "step": 240
+    },
+    {
+      "clip_ratio": 0.0013090946631564293,
+      "completion_length": 792.5536079406738,
+      "epoch": 0.03596847161979385,
+      "grad_norm": 0.047966547310352325,
+      "kl": 0.03855133056640625,
+      "learning_rate": 2.9900200099795396e-06,
+      "loss": 0.0004,
+      "num_tokens": 12813969.0,
+      "reward": 0.5236398167908192,
+      "reward_std": 0.19648715062066913,
+      "rewards/code_reward": 0.42238980112597346,
+      "rewards/format_reward": 1.0124999918043613,
+      "step": 241
+    },
+    {
+      "clip_ratio": 0.0011606769148784224,
+      "epoch": 0.03611771839000047,
+      "grad_norm": 0.050562020391225815,
+      "kl": 0.038921356201171875,
+      "learning_rate": 2.9755238402607826e-06,
+      "loss": 0.0005,
+      "step": 242
+    },
+    {
+      "clip_ratio": 0.0010873440842260607,
+      "epoch": 0.03626696516020708,
+      "grad_norm": 0.046209290623664856,
+      "kl": 0.03893280029296875,
+      "learning_rate": 2.961018208013367e-06,
+      "loss": 0.0003,
+      "step": 243
+    },
+    {
+      "clip_ratio": 0.0011430384001869243,
+      "completion_length": 645.9375267028809,
+      "epoch": 0.036416211930413694,
+      "grad_norm": 0.04914650693535805,
+      "kl": 0.046295166015625,
+      "learning_rate": 2.9465037218645694e-06,
+      "loss": 0.0005,
+      "num_tokens": 12937253.0,
+      "reward": 0.6496362583711743,
+      "reward_std": 0.18762771505862474,
+      "rewards/code_reward": 0.5519576612859964,
+      "rewards/format_reward": 0.9767857156693935,
+      "step": 244
+    },
+    {
+      "clip_ratio": 0.0009474948765273439,
+      "epoch": 0.03656545870062031,
+      "grad_norm": 0.050520021468400955,
+      "kl": 0.0462493896484375,
+      "learning_rate": 2.9319809908131604e-06,
+      "loss": 0.0003,
+      "step": 245
+    },
+    {
+      "clip_ratio": 0.0009283031004088116,
+      "epoch": 0.03671470547082692,
+      "grad_norm": 0.04778316617012024,
+      "kl": 0.04526519775390625,
+      "learning_rate": 2.917450624203847e-06,
+      "loss": 0.0002,
+      "step": 246
+    },
+    {
+      "clip_ratio": 0.0011882597391377203,
+      "completion_length": 697.1339626312256,
+      "epoch": 0.03686395224103353,
+      "grad_norm": 0.04777618125081062,
+      "kl": 0.041156768798828125,
+      "learning_rate": 2.9029132317017118e-06,
+      "loss": 0.0004,
+      "num_tokens": 13075518.0,
+      "reward": 0.6029732944443822,
+      "reward_std": 0.20219829003326595,
+      "rewards/code_reward": 0.5043125713709742,
+      "rewards/format_reward": 0.9866071343421936,
+      "step": 247
+    },
+    {
+      "clip_ratio": 0.0011061100340157282,
+      "epoch": 0.037013199011240144,
+      "grad_norm": 0.046102263033390045,
+      "kl": 0.042461395263671875,
+      "learning_rate": 2.888369423266629e-06,
+      "loss": 0.0003,
+      "step": 248
+    },
+    {
+      "clip_ratio": 0.0012012151892122347,
+      "epoch": 0.037162445781446764,
+      "grad_norm": 0.049948278814554214,
+      "kl": 0.041042327880859375,
+      "learning_rate": 2.8738198091276712e-06,
+      "loss": 0.0003,
+      "step": 249
+    },
+    {
+      "clip_ratio": 0.0011589178429858293,
+      "completion_length": 880.8750400543213,
+      "epoch": 0.037311692551653376,
+      "grad_norm": 0.046810079365968704,
+      "kl": 0.0442352294921875,
+      "learning_rate": 2.859264999757509e-06,
+      "loss": 0.0005,
+      "num_tokens": 13245436.0,
+      "reward": 0.47556353313848376,
+      "reward_std": 0.2117939020972699,
+      "rewards/code_reward": 0.3802063832990825,
+      "rewards/format_reward": 0.9535714266821742,
+      "step": 250
+    },
+    {
+      "clip_ratio": 0.001155683581600897,
+      "epoch": 0.03746093932185999,
+      "grad_norm": 0.04762619733810425,
+      "kl": 0.04388427734375,
+      "learning_rate": 2.8447056058467928e-06,
+      "loss": 0.0004,
+      "step": 251
+    },
+    {
+      "clip_ratio": 0.0012009528154521831,
+      "epoch": 0.0376101860920666,
+      "grad_norm": 0.046324875205755234,
+      "kl": 0.0436859130859375,
+      "learning_rate": 2.830142238278531e-06,
+      "loss": 0.0004,
+      "step": 252
+    },
+    {
+      "clip_ratio": 0.0011499394786369521,
+      "completion_length": 702.1071796417236,
+      "epoch": 0.037759432862273214,
+      "grad_norm": 0.05207349359989166,
+      "kl": 0.047008514404296875,
+      "learning_rate": 2.81557550810246e-06,
+      "loss": 0.0007,
+      "num_tokens": 13379036.0,
+      "reward": 0.5698134193662554,
+      "reward_std": 0.18162817368283868,
+      "rewards/code_reward": 0.47561697755008936,
+      "rewards/format_reward": 0.9419642891734838,
+      "step": 253
+    },
+    {
+      "clip_ratio": 0.0009552986102789873,
+      "epoch": 0.037908679632479826,
+      "grad_norm": 0.05312632396817207,
+      "kl": 0.0461883544921875,
+      "learning_rate": 2.8010060265094026e-06,
+      "loss": 0.0007,
+      "step": 254
+    },
+    {
+      "clip_ratio": 0.0012589068483066512,
+      "epoch": 0.03805792640268644,
+      "grad_norm": 0.05023624747991562,
+      "kl": 0.045841217041015625,
+      "learning_rate": 2.786434404805629e-06,
+      "loss": 0.0004,
+      "step": 255
+    },
+    {
+      "clip_ratio": 0.0014080564960750053,
+      "completion_length": 672.3393173217773,
+      "epoch": 0.03820717317289306,
+      "grad_norm": 0.0485769547522068,
+      "kl": 0.04241943359375,
+      "learning_rate": 2.771861254387199e-06,
+      "loss": 0.0005,
+      "num_tokens": 13512438.0,
+      "reward": 0.5117118041962385,
+      "reward_std": 0.23872398328967392,
+      "rewards/code_reward": 0.41510464530438185,
+      "rewards/format_reward": 0.966071430593729,
+      "step": 256
+    },
+    {
+      "clip_ratio": 0.0011822391352325212,
+      "epoch": 0.03835641994309967,
+      "grad_norm": 0.050369687378406525,
+      "kl": 0.04282379150390625,
+      "learning_rate": 2.7572871867143204e-06,
+      "loss": 0.0004,
+      "step": 257
+    },
+    {
+      "clip_ratio": 0.0014788927746849367,
+      "epoch": 0.03850566671330628,
+      "grad_norm": 0.04701628535985947,
+      "kl": 0.04404449462890625,
+      "learning_rate": 2.742712813285681e-06,
+      "loss": 0.0004,
+      "step": 258
+    },
+    {
+      "clip_ratio": 0.0010453241848153993,
+      "completion_length": 773.8303928375244,
+      "epoch": 0.038654913483512895,
+      "grad_norm": 0.04721568897366524,
+      "kl": 0.039340972900390625,
+      "learning_rate": 2.7281387456128017e-06,
+      "loss": 0.0004,
+      "num_tokens": 13657739.0,
+      "reward": 0.5111014764988795,
+      "reward_std": 0.20084915030747652,
+      "rewards/code_reward": 0.4129764676472405,
+      "rewards/format_reward": 0.9812499936670065,
+      "step": 259
+    },
+    {
+      "clip_ratio": 0.001074462946235144,
+      "epoch": 0.03880416025371951,
+      "grad_norm": 0.04527387022972107,
+      "kl": 0.0382232666015625,
+      "learning_rate": 2.7135655951943716e-06,
+      "loss": 0.0005,
+      "step": 260
+    },
+    {
+      "clip_ratio": 0.0010467221291037276,
+      "epoch": 0.03895340702392612,
+      "grad_norm": 0.04464754834771156,
+      "kl": 0.036327362060546875,
+      "learning_rate": 2.698993973490598e-06,
+      "loss": 0.0004,
+      "step": 261
+    },
+    {
+      "clip_ratio": 0.0011872881314047845,
+      "completion_length": 739.3036079406738,
+      "epoch": 0.03910265379413273,
+      "grad_norm": 0.04908748343586922,
+      "kl": 0.056262969970703125,
+      "learning_rate": 2.6844244918975416e-06,
+      "loss": 0.0005,
+      "num_tokens": 13804031.0,
+      "reward": 0.6070374413393438,
+      "reward_std": 0.23497100081294775,
+      "rewards/code_reward": 0.5173052740865387,
+      "rewards/format_reward": 0.897321430966258,
+      "step": 262
+    },
+    {
+      "clip_ratio": 0.001177058060420677,
+      "epoch": 0.03925190056433935,
+      "grad_norm": 0.054567039012908936,
+      "kl": 0.054866790771484375,
+      "learning_rate": 2.66985776172147e-06,
+      "loss": 0.0005,
+      "step": 263
+    },
+    {
+      "clip_ratio": 0.00118064760681591,
+      "epoch": 0.039401147334545965,
+      "grad_norm": 0.042729251086711884,
+      "kl": 0.047893524169921875,
+      "learning_rate": 2.6552943941532088e-06,
+      "loss": 0.0004,
+      "step": 264
+    },
+    {
+      "clip_ratio": 0.0013072448928141966,
+      "completion_length": 849.7143383026123,
+      "epoch": 0.03955039410475258,
+      "grad_norm": 0.04685011878609657,
+      "kl": 0.03894233703613281,
+      "learning_rate": 2.6407350002424927e-06,
+      "loss": 0.0005,
+      "num_tokens": 13971159.0,
+      "reward": 0.5141286924481392,
+      "reward_std": 0.27967278216965497,
+      "rewards/code_reward": 0.415021532215178,
+      "rewards/format_reward": 0.991071417927742,
+      "step": 265
+    },
+    {
+      "clip_ratio": 0.001456160111047211,
+      "epoch": 0.03969964087495919,
+      "grad_norm": 0.0907537117600441,
+      "kl": 0.036434173583984375,
+      "learning_rate": 2.626180190872329e-06,
+      "loss": 0.0005,
+      "step": 266
+    },
+    {
+      "clip_ratio": 0.0013232155906734988,
+      "epoch": 0.0398488876451658,
+      "grad_norm": 0.05118626356124878,
+      "kl": 0.03912162780761719,
+      "learning_rate": 2.611630576733372e-06,
+      "loss": 0.0004,
+      "step": 267
+    },
+    {
+      "clip_ratio": 0.001450759154977277,
+      "completion_length": 746.2321796417236,
+      "epoch": 0.039998134415372415,
+      "grad_norm": 0.05404108017683029,
+      "kl": 0.039783477783203125,
+      "learning_rate": 2.5970867682982885e-06,
+      "loss": 0.0005,
+      "num_tokens": 14119900.0,
+      "reward": 0.426578835118562,
+      "reward_std": 0.2707485407590866,
+      "rewards/code_reward": 0.33131096488796175,
+      "rewards/format_reward": 0.9526785761117935,
+      "step": 268
+    },
+    {
+      "clip_ratio": 0.001243086520844372,
+      "epoch": 0.040147381185579034,
+      "grad_norm": 0.04573540762066841,
+      "kl": 0.040866851806640625,
+      "learning_rate": 2.582549375796154e-06,
+      "loss": 0.0005,
+      "step": 269
+    },
+    {
+      "clip_ratio": 0.001273469748412026,
+      "epoch": 0.04029662795578565,
+      "grad_norm": 0.04555228725075722,
+      "kl": 0.041530609130859375,
+      "learning_rate": 2.568019009186841e-06,
+      "loss": 0.0004,
+      "step": 270
+    },
+    {
+      "clip_ratio": 0.0010037624397227773,
+      "completion_length": 683.7321643829346,
+      "epoch": 0.04044587472599226,
+      "grad_norm": 0.04558391869068146,
+      "kl": 0.035755157470703125,
+      "learning_rate": 2.5534962781354317e-06,
+      "loss": 0.0005,
+      "num_tokens": 14254592.0,
+      "reward": 0.7018171134404838,
+      "reward_std": 0.1548383531626314,
+      "rewards/code_reward": 0.6096742328445544,
+      "rewards/format_reward": 0.9214285779744387,
+      "step": 271
+    },
+    {
+      "clip_ratio": 0.0009475577735429397,
+      "epoch": 0.04059512149619887,
+      "grad_norm": 0.04442001134157181,
+      "kl": 0.03681182861328125,
+      "learning_rate": 2.538981791986634e-06,
+      "loss": 0.0004,
+      "step": 272
+    },
+    {
+      "clip_ratio": 0.0009967278165277094,
+      "epoch": 0.040744368266405484,
+      "grad_norm": 0.04307863861322403,
+      "kl": 0.03729248046875,
+      "learning_rate": 2.524476159739218e-06,
+      "loss": 0.0004,
+      "step": 273
+    },
+    {
+      "clip_ratio": 0.0009341392360511236,
+      "completion_length": 763.7232418060303,
+      "epoch": 0.040893615036612097,
+      "grad_norm": 0.04648442193865776,
+      "kl": 0.036266326904296875,
+      "learning_rate": 2.5099799900204607e-06,
+      "loss": 0.0004,
+      "num_tokens": 14407973.0,
+      "reward": 0.5817076284438372,
+      "reward_std": 0.19802262401208282,
+      "rewards/code_reward": 0.48893973045778694,
+      "rewards/format_reward": 0.9276785720139742,
+      "step": 274
+    },
+    {
+      "clip_ratio": 0.0009563791154505452,
+      "epoch": 0.04104286180681871,
+      "grad_norm": 0.043009962886571884,
+      "kl": 0.03775787353515625,
+      "learning_rate": 2.4954938910606108e-06,
+      "loss": 0.0003,
+      "step": 275
+    },
+    {
+      "clip_ratio": 0.0009447472584724892,
+      "epoch": 0.04119210857702533,
+      "grad_norm": 0.044043902307748795,
+      "kl": 0.03899383544921875,
+      "learning_rate": 2.481018470667368e-06,
+      "loss": 0.0003,
+      "step": 276
+    },
+    {
+      "clip_ratio": 0.0009749470536917215,
+      "completion_length": 751.8839569091797,
+      "epoch": 0.04134135534723194,
+      "grad_norm": 0.04722112417221069,
+      "kl": 0.051502227783203125,
+      "learning_rate": 2.4665543362003802e-06,
+      "loss": 0.0005,
+      "num_tokens": 14552930.0,
+      "reward": 0.463415396399796,
+      "reward_std": 0.1340479573700577,
+      "rewards/code_reward": 0.3710046644337126,
+      "rewards/format_reward": 0.9241071259602904,
+      "step": 277
+    },
+    {
+      "clip_ratio": 0.0009320784265582915,
+      "epoch": 0.04149060211743855,
+      "grad_norm": 0.047061435878276825,
+      "kl": 0.0539093017578125,
+      "learning_rate": 2.4521020945457615e-06,
+      "loss": 0.0005,
+      "step": 278
+    },
+    {
+      "clip_ratio": 0.0011352980673109414,
+      "epoch": 0.041639848887645166,
+      "grad_norm": 0.0768757089972496,
+      "kl": 0.0503387451171875,
+      "learning_rate": 2.4376623520906255e-06,
+      "loss": 0.0004,
+      "step": 279
+    },
+    {
+      "clip_ratio": 0.0011972093489021063,
+      "completion_length": 795.5089569091797,
+      "epoch": 0.04178909565785178,
+      "grad_norm": 0.04606209695339203,
+      "kl": 0.06067657470703125,
+      "learning_rate": 2.4232357146976478e-06,
+      "loss": 0.0006,
+      "num_tokens": 14708779.0,
+      "reward": 0.6148986229673028,
+      "reward_std": 0.18130915402434766,
+      "rewards/code_reward": 0.5234700404689647,
+      "rewards/format_reward": 0.9142857044935226,
+      "step": 280
+    },
+    {
+      "clip_ratio": 0.0010662599033821607,
+      "epoch": 0.04193834242805839,
+      "grad_norm": 0.04503650590777397,
+      "kl": 0.0595855712890625,
+      "learning_rate": 2.408822787679637e-06,
+      "loss": 0.0006,
+      "step": 281
+    },
+    {
+      "clip_ratio": 0.0012542046988528455,
+      "epoch": 0.042087589198265,
+      "grad_norm": 0.04570099711418152,
+      "kl": 0.054462432861328125,
+      "learning_rate": 2.3944241757741475e-06,
+      "loss": 0.0005,
+      "step": 282
+    },
+    {
+      "clip_ratio": 0.0010910165510722436,
+      "completion_length": 841.4821853637695,
+      "epoch": 0.04223683596847162,
+      "grad_norm": 0.043202780187129974,
+      "kl": 0.04636383056640625,
+      "learning_rate": 2.380040483118097e-06,
+      "loss": 0.0005,
+      "num_tokens": 14872542.0,
+      "reward": 0.47422279929742217,
+      "reward_std": 0.21761185978539288,
+      "rewards/code_reward": 0.3716334954369813,
+      "rewards/format_reward": 1.0258928649127483,
+      "step": 283
+    },
+    {
+      "clip_ratio": 0.0011495619401102886,
+      "epoch": 0.042386082738678235,
+      "grad_norm": 0.042755380272865295,
+      "kl": 0.044036865234375,
+      "learning_rate": 2.365672313222419e-06,
+      "loss": 0.0004,
+      "step": 284
+    },
+    {
+      "clip_ratio": 0.0013028225057496456,
+      "epoch": 0.04253532950888485,
+      "grad_norm": 0.05651061609387398,
+      "kl": 0.042083740234375,
+      "learning_rate": 2.351320268946749e-06,
+      "loss": 0.0005,
+      "step": 285
+    },
+    {
+      "clip_ratio": 0.001140464471973246,
+      "completion_length": 1019.5000534057617,
+      "epoch": 0.04268457627909146,
+      "grad_norm": 0.05123982951045036,
+      "kl": 0.05125236511230469,
+      "learning_rate": 2.336984952474119e-06,
+      "loss": 0.0005,
+      "num_tokens": 15060254.0,
+      "reward": 0.48235001834109426,
+      "reward_std": 0.1879777810536325,
+      "rewards/code_reward": 0.38779641879955307,
+      "rewards/format_reward": 0.9455357119441032,
+      "step": 286
+    },
+    {
+      "clip_ratio": 0.0010813200133270584,
+      "epoch": 0.04283382304929807,
+      "grad_norm": 0.049375370144844055,
+      "kl": 0.047840118408203125,
+      "learning_rate": 2.322666965285697e-06,
+      "loss": 0.0004,
+      "step": 287
+    },
+    {
+      "clip_ratio": 0.0010197247520409292,
+      "epoch": 0.042983069819504685,
+      "grad_norm": 0.047002602368593216,
+      "kl": 0.04539680480957031,
+      "learning_rate": 2.3083669081355507e-06,
+      "loss": 0.0004,
+      "step": 288
+    },
+    {
+      "clip_ratio": 0.0010922325645879027,
+      "completion_length": 860.9821758270264,
+      "epoch": 0.0431323165897113,
+      "grad_norm": 0.047084398567676544,
+      "kl": 0.033111572265625,
+      "learning_rate": 2.2940853810254377e-06,
+      "loss": 0.0003,
+      "num_tokens": 15215862.0,
+      "reward": 0.41257912619039416,
+      "reward_std": 0.17679887195117772,
+      "rewards/code_reward": 0.30829338620242197,
+      "rewards/format_reward": 1.0428571440279484,
+      "step": 289
+    },
+    {
+      "clip_ratio": 0.0011078603311034385,
+      "epoch": 0.04328156335991792,
+      "grad_norm": 0.047457680106163025,
+      "kl": 0.032649993896484375,
+      "learning_rate": 2.2798229831796313e-06,
+      "loss": 0.0004,
+      "step": 290
+    },
+    {
+      "clip_ratio": 0.0012763461973008816,
+      "epoch": 0.04343081013012453,
+      "grad_norm": 0.04568823054432869,
+      "kl": 0.032649993896484375,
+      "learning_rate": 2.2655803130197816e-06,
+      "loss": 0.0003,
+      "step": 291
+    },
+    {
+      "clip_ratio": 0.0011839414237329038,
+      "completion_length": 830.1964721679688,
+      "epoch": 0.04358005690033114,
+      "grad_norm": 0.04152772203087807,
+      "kl": 0.038677215576171875,
+      "learning_rate": 2.2513579681398034e-06,
+      "loss": 0.0005,
+      "num_tokens": 15367448.0,
+      "reward": 0.5385435288771987,
+      "reward_std": 0.19832654343917966,
+      "rewards/code_reward": 0.4406863655894995,
+      "rewards/format_reward": 0.9785714186728001,
+      "step": 292
+    },
+    {
+      "clip_ratio": 0.0010050950450022356,
+      "epoch": 0.043729303670537754,
+      "grad_norm": 0.04242389649152756,
+      "kl": 0.03912353515625,
+      "learning_rate": 2.237156545280803e-06,
+      "loss": 0.0004,
+      "step": 293
+    },
+    {
+      "clip_ratio": 0.001050207116350066,
+      "epoch": 0.04387855044074437,
+      "grad_norm": 0.04236888512969017,
+      "kl": 0.039356231689453125,
+      "learning_rate": 2.2229766403060403e-06,
+      "loss": 0.0004,
+      "step": 294
+    },
+    {
+      "clip_ratio": 0.0011052291774831247,
+      "completion_length": 796.7411079406738,
+      "epoch": 0.04402779721095098,
+      "grad_norm": 0.062591552734375,
+      "kl": 0.059398651123046875,
+      "learning_rate": 2.2088188481759305e-06,
+      "loss": 0.0006,
+      "num_tokens": 15521069.0,
+      "reward": 0.7143844980746508,
+      "reward_std": 0.22826048522256315,
+      "rewards/code_reward": 0.6077773505821824,
+      "rewards/format_reward": 1.066071417182684,
+      "step": 295
+    },
+    {
+      "clip_ratio": 0.001126413505517121,
+      "epoch": 0.04417704398115759,
+      "grad_norm": 0.05377168208360672,
+      "kl": 0.052120208740234375,
+      "learning_rate": 2.194683762923073e-06,
+      "loss": 0.0006,
+      "step": 296
+    },
+    {
+      "clip_ratio": 0.0010959034188999794,
+      "epoch": 0.04432629075136421,
+      "grad_norm": 0.049911994487047195,
+      "kl": 0.0478057861328125,
+      "learning_rate": 2.1805719776273387e-06,
+      "loss": 0.0005,
+      "step": 297
+    },
+    {
+      "clip_ratio": 0.0011420191494835308,
+      "completion_length": 801.7411098480225,
+      "epoch": 0.044475537521570824,
+      "grad_norm": 0.04711169749498367,
+      "kl": 0.047809600830078125,
+      "learning_rate": 2.166484084390974e-06,
+      "loss": 0.0005,
+      "num_tokens": 15678309.0,
+      "reward": 0.3837883761152625,
+      "reward_std": 0.14018947025761008,
+      "rewards/code_reward": 0.27164551382884383,
+      "rewards/format_reward": 1.1214285604655743,
+      "step": 298
+    },
+    {
+      "clip_ratio": 0.0009830774752117577,
+      "epoch": 0.044624784291777436,
+      "grad_norm": 0.04708566516637802,
+      "kl": 0.047992706298828125,
+      "learning_rate": 2.1524206743137636e-06,
+      "loss": 0.0004,
+      "step": 299
+    },
+    {
+      "clip_ratio": 0.000958900871410151,
+      "epoch": 0.04477403106198405,
+      "grad_norm": 0.045896608382463455,
+      "kl": 0.049419403076171875,
+      "learning_rate": 2.1383823374682287e-06,
+      "loss": 0.0004,
+      "step": 300
+    },
+    {
+      "clip_ratio": 0.0012122261905460618,
+      "completion_length": 859.7143230438232,
+      "epoch": 0.04492327783219066,
+      "grad_norm": 0.04249017313122749,
+      "kl": 0.044986724853515625,
+      "learning_rate": 2.124369662874868e-06,
+      "loss": 0.0004,
+      "num_tokens": 15841559.0,
+      "reward": 0.548895129468292,
+      "reward_std": 0.24452911573462188,
+      "rewards/code_reward": 0.4446094058221206,
+      "rewards/format_reward": 1.0428571291267872,
+      "step": 301
+    },
+    {
+      "clip_ratio": 0.001091581678338116,
+      "epoch": 0.045072524602397274,
+      "grad_norm": 0.04369416460394859,
+      "kl": 0.044300079345703125,
+      "learning_rate": 2.110383238477441e-06,
+      "loss": 0.0005,
+      "step": 302
+    },
+    {
+      "clip_ratio": 0.0011382698903616983,
+      "epoch": 0.04522177137260389,
+      "grad_norm": 0.041453346610069275,
+      "kl": 0.04480743408203125,
+      "learning_rate": 2.096423651118305e-06,
+      "loss": 0.0004,
+      "step": 303
+    },
+    {
+      "clip_ratio": 0.0008476753027935047,
+      "completion_length": 781.8482494354248,
+      "epoch": 0.045371018142810506,
+      "grad_norm": 0.045421402901411057,
+      "kl": 0.04686737060546875,
+      "learning_rate": 2.082491486513788e-06,
+      "loss": 0.0005,
+      "num_tokens": 15989375.0,
+      "reward": 0.6513705789111555,
+      "reward_std": 0.23313539451919496,
+      "rewards/code_reward": 0.5432455567643046,
+      "rewards/format_reward": 1.0812499970197678,
+      "step": 304
+    },
+    {
+      "clip_ratio": 0.0011021571021956333,
+      "epoch": 0.04552026491301712,
+      "grad_norm": 0.04760179668664932,
+      "kl": 0.045501708984375,
+      "learning_rate": 2.0685873292296116e-06,
+      "loss": 0.0005,
+      "step": 305
+    },
+    {
+      "clip_ratio": 0.0009927952487487346,
+      "epoch": 0.04566951168322373,
+      "grad_norm": 0.04543590918183327,
+      "kl": 0.043498992919921875,
+      "learning_rate": 2.054711762656369e-06,
+      "loss": 0.0004,
+      "step": 306
+    },
+    {
+      "clip_ratio": 0.001191185787320137,
+      "completion_length": 772.3750400543213,
+      "epoch": 0.04581875845343034,
+      "grad_norm": 0.04691977798938751,
+      "kl": 0.03873443603515625,
+      "learning_rate": 2.040865368985044e-06,
+      "loss": 0.0004,
+      "num_tokens": 16132539.0,
+      "reward": 0.5620117126964033,
+      "reward_std": 0.23455260833725333,
+      "rewards/code_reward": 0.4617438427376328,
+      "rewards/format_reward": 1.0026785656809807,
+      "step": 307
+    },
+    {
+      "clip_ratio": 0.0011454644645709777,
+      "epoch": 0.045968005223636955,
+      "grad_norm": 0.045798543840646744,
+      "kl": 0.03643035888671875,
+      "learning_rate": 2.027048729182583e-06,
+      "loss": 0.0005,
+      "step": 308
+    },
+    {
+      "clip_ratio": 0.0013067667568975594,
+      "epoch": 0.04611725199384357,
+      "grad_norm": 0.0446229986846447,
+      "kl": 0.03519439697265625,
+      "learning_rate": 2.0132624229675205e-06,
+      "loss": 0.0003,
+      "step": 309
+    },
+    {
+      "clip_ratio": 0.0013392006876529194,
+      "completion_length": 795.8036117553711,
+      "epoch": 0.04626649876405019,
+      "grad_norm": 0.044981542974710464,
+      "kl": 0.045459747314453125,
+      "learning_rate": 1.9995070287856546e-06,
+      "loss": 0.0004,
+      "num_tokens": 16287462.0,
+      "reward": 0.5069911074824631,
+      "reward_std": 0.2646176088601351,
+      "rewards/code_reward": 0.40404467517510056,
+      "rewards/format_reward": 1.0294642690569162,
+      "step": 310
+    },
+    {
+      "clip_ratio": 0.0012751408075928339,
+      "epoch": 0.0464157455342568,
+      "grad_norm": 0.04428853467106819,
+      "kl": 0.04509735107421875,
+      "learning_rate": 1.985783123785774e-06,
+      "loss": 0.0004,
+      "step": 311
+    },
+    {
+      "clip_ratio": 0.001256519486560137,
+      "epoch": 0.04656499230446341,
+      "grad_norm": 0.04464598372578621,
+      "kl": 0.0438995361328125,
+      "learning_rate": 1.9720912837954486e-06,
+      "loss": 0.0004,
+      "step": 312
+    },
+    {
+      "clip_ratio": 0.0010024435077866656,
+      "completion_length": 737.3661136627197,
+      "epoch": 0.046714239074670025,
+      "grad_norm": 0.0473940409719944,
+      "kl": 0.0488128662109375,
+      "learning_rate": 1.958432083296862e-06,
+      "loss": 0.0007,
+      "num_tokens": 16426698.0,
+      "reward": 0.577216680161655,
+      "reward_std": 0.24476247746497393,
+      "rewards/code_reward": 0.47391308448277414,
+      "rewards/format_reward": 1.033035695552826,
+      "step": 313
+    },
+    {
+      "clip_ratio": 0.0009519284703856101,
+      "epoch": 0.04686348584487664,
+      "grad_norm": 0.046678487211465836,
+      "kl": 0.04576873779296875,
+      "learning_rate": 1.9448060954027093e-06,
+      "loss": 0.0005,
+      "step": 314
+    },
+    {
+      "clip_ratio": 0.0010747394862846704,
+      "epoch": 0.04701273261508325,
+      "grad_norm": 0.05263395234942436,
+      "kl": 0.04498291015625,
+      "learning_rate": 1.931213891832153e-06,
+      "loss": 0.0006,
+      "step": 315
+    },
+    {
+      "clip_ratio": 0.000990249405731447,
+      "completion_length": 877.2232570648193,
+      "epoch": 0.04716197938528986,
+      "grad_norm": 0.07013804465532303,
+      "kl": 0.0428314208984375,
+      "learning_rate": 1.9176560428868336e-06,
+      "loss": 0.0005,
+      "num_tokens": 16593190.0,
+      "reward": 0.5014848634600639,
+      "reward_std": 0.1569962182547897,
+      "rewards/code_reward": 0.3947884114459157,
+      "rewards/format_reward": 1.0669642873108387,
+      "step": 316
+    },
+    {
+      "clip_ratio": 0.001294584584684344,
+      "epoch": 0.04731122615549648,
+      "grad_norm": 0.04042747989296913,
+      "kl": 0.042980194091796875,
+      "learning_rate": 1.9041331174269373e-06,
+      "loss": 0.0006,
+      "step": 317
+    },
+    {
+      "clip_ratio": 0.00103904002935451,
+      "epoch": 0.047460472925703094,
+      "grad_norm": 0.04211616516113281,
+      "kl": 0.0416107177734375,
+      "learning_rate": 1.8906456828473341e-06,
+      "loss": 0.0005,
+      "step": 318
+    },
+    {
+      "clip_ratio": 0.0009875956375253736,
+      "completion_length": 904.2768287658691,
+      "epoch": 0.04760971969590971,
+      "grad_norm": 0.05007445812225342,
+      "kl": 0.040981292724609375,
+      "learning_rate": 1.8771943050537656e-06,
+      "loss": 0.0003,
+      "num_tokens": 16763902.0,
+      "reward": 0.3916885075159371,
+      "reward_std": 0.23931702435947955,
+      "rewards/code_reward": 0.30088492622599006,
+      "rewards/format_reward": 0.9080357123166323,
+      "step": 319
+    },
+    {
+      "clip_ratio": 0.0008391569135710597,
+      "epoch": 0.04775896646611632,
+      "grad_norm": 0.05043473839759827,
+      "kl": 0.042842864990234375,
+      "learning_rate": 1.8637795484391046e-06,
+      "loss": 0.0004,
+      "step": 320
+    },
+    {
+      "clip_ratio": 0.0010355508493375964,
+      "epoch": 0.04790821323632293,
+      "grad_norm": 0.34376153349876404,
+      "kl": 0.0391998291015625,
+      "learning_rate": 1.8504019758596698e-06,
+      "loss": 0.0004,
+      "step": 321
+    },
+    {
+      "clip_ratio": 0.0010621717938192887,
+      "completion_length": 850.4196853637695,
+      "epoch": 0.048057460006529544,
+      "grad_norm": 0.04401549696922302,
+      "kl": 0.035259246826171875,
+      "learning_rate": 1.8370621486116163e-06,
+      "loss": 0.0003,
+      "num_tokens": 16924081.0,
+      "reward": 0.4363535214215517,
+      "reward_std": 0.18411048781126738,
+      "rewards/code_reward": 0.34376421920751454,
+      "rewards/format_reward": 0.9258928671479225,
+      "step": 322
+    },
+    {
+      "clip_ratio": 0.001003089109872235,
+      "epoch": 0.048206706776736157,
+      "grad_norm": 0.04463668912649155,
+      "kl": 0.036479949951171875,
+      "learning_rate": 1.823760626407377e-06,
+      "loss": 0.0003,
+      "step": 323
+    },
+    {
+      "clip_ratio": 0.0010362897501181578,
+      "epoch": 0.048355953546942776,
+      "grad_norm": 0.0432155467569828,
+      "kl": 0.036895751953125,
+      "learning_rate": 1.8104979673521838e-06,
+      "loss": 0.0003,
+      "step": 324
+    },
+    {
+      "clip_ratio": 0.001312947628321126,
+      "completion_length": 796.1696815490723,
+      "epoch": 0.04850520031714939,
+      "grad_norm": 0.04366760700941086,
+      "kl": 0.03292655944824219,
+      "learning_rate": 1.7972747279206482e-06,
+      "loss": 0.0004,
+      "num_tokens": 17068475.0,
+      "reward": 0.6358563541434705,
+      "reward_std": 0.2135081202723086,
+      "rewards/code_reward": 0.5282670482993126,
+      "rewards/format_reward": 1.075892847031355,
+      "step": 325
+    },
+    {
+      "clip_ratio": 0.0013257294258437469,
+      "epoch": 0.048654447087356,
+      "grad_norm": 0.04495041444897652,
+      "kl": 0.03363227844238281,
+      "learning_rate": 1.7840914629334122e-06,
+      "loss": 0.0004,
+      "step": 326
+    },
+    {
+      "clip_ratio": 0.0012908969647469348,
+      "epoch": 0.04880369385756261,
+      "grad_norm": 0.0467698760330677,
+      "kl": 0.03423309326171875,
+      "learning_rate": 1.7709487255338731e-06,
+      "loss": 0.0003,
+      "step": 327
+    },
+    {
+      "clip_ratio": 0.0012681423613685183,
+      "completion_length": 673.5893173217773,
+      "epoch": 0.048952940627769226,
+      "grad_norm": 0.046931758522987366,
+      "kl": 0.037639617919921875,
+      "learning_rate": 1.7578470671649684e-06,
+      "loss": 0.0003,
+      "num_tokens": 17198167.0,
+      "reward": 0.6159717463888228,
+      "reward_std": 0.24445109069347382,
+      "rewards/code_reward": 0.4978467100299895,
+      "rewards/format_reward": 1.1812499985098839,
+      "step": 328
+    },
+    {
+      "clip_ratio": 0.0013088965934002772,
+      "epoch": 0.04910218739797584,
+      "grad_norm": 0.04621303454041481,
+      "kl": 0.039310455322265625,
+      "learning_rate": 1.744787037546045e-06,
+      "loss": 0.0004,
+      "step": 329
+    },
+    {
+      "clip_ratio": 0.001249867600563448,
+      "epoch": 0.04925143416818245,
+      "grad_norm": 0.04626749828457832,
+      "kl": 0.03778076171875,
+      "learning_rate": 1.731769184649788e-06,
+      "loss": 0.0003,
+      "step": 330
+    },
+    {
+      "clip_ratio": 0.0011540742816578131,
+      "completion_length": 870.2857570648193,
+      "epoch": 0.04940068093838907,
+      "grad_norm": 0.04759405553340912,
+      "kl": 0.030445098876953125,
+      "learning_rate": 1.7187940546792325e-06,
+      "loss": 0.0003,
+      "num_tokens": 17360529.0,
+      "reward": 0.3948647300712764,
+      "reward_std": 0.234670925186947,
+      "rewards/code_reward": 0.3062932917528087,
+      "rewards/format_reward": 0.8857142850756645,
+      "step": 331
+    },
+    {
+      "clip_ratio": 0.0012074910209776135,
+      "epoch": 0.04954992770859568,
+      "grad_norm": 0.044540390372276306,
+      "kl": 0.03099822998046875,
+      "learning_rate": 1.7058621920448465e-06,
+      "loss": 0.0003,
+      "step": 332
+    },
+    {
+      "clip_ratio": 0.0011028643830286455,
+      "epoch": 0.049699174478802295,
+      "grad_norm": 0.05389461666345596,
+      "kl": 0.031070709228515625,
+      "learning_rate": 1.6929741393416855e-06,
+      "loss": 0.0001,
+      "step": 333
+    },
+    {
+      "clip_ratio": 0.0012008915800834075,
+      "completion_length": 834.9464569091797,
+      "epoch": 0.04984842124900891,
+      "grad_norm": 0.043137501925230026,
+      "kl": 0.029438018798828125,
+      "learning_rate": 1.6801304373266286e-06,
+      "loss": 0.0004,
+      "num_tokens": 17520984.0,
+      "reward": 0.6375483605079353,
+      "reward_std": 0.24400542792864144,
+      "rewards/code_reward": 0.5356733408989385,
+      "rewards/format_reward": 1.0187499970197678,
+      "step": 334
+    },
+    {
+      "clip_ratio": 0.0010621378514770186,
+      "epoch": 0.04999766801921552,
+      "grad_norm": 0.04156143590807915,
+      "kl": 0.030574798583984375,
+      "learning_rate": 1.667331624895689e-06,
+      "loss": 0.0004,
+      "step": 335
+    },
+    {
+      "clip_ratio": 0.0010844554426512332,
+      "epoch": 0.05014691478942213,
+      "grad_norm": 0.04887258633971214,
+      "kl": 0.029483795166015625,
+      "learning_rate": 1.6545782390614037e-06,
+      "loss": 0.0003,
+      "step": 336
+    },
+    {
+      "clip_ratio": 0.0011277581979811657,
+      "completion_length": 799.4732551574707,
+      "epoch": 0.05029616155962875,
+      "grad_norm": 0.04529580846428871,
+      "kl": 0.03227424621582031,
+      "learning_rate": 1.6418708149302992e-06,
+      "loss": 0.0005,
+      "num_tokens": 17677928.0,
+      "reward": 0.6606419142335653,
+      "reward_std": 0.2389362787362188,
+      "rewards/code_reward": 0.5465347599820234,
+      "rewards/format_reward": 1.1410714201629162,
+      "step": 337
+    },
+    {
+      "clip_ratio": 0.0009562958011883893,
+      "epoch": 0.050445408329835364,
+      "grad_norm": 0.0461302287876606,
+      "kl": 0.032329559326171875,
+      "learning_rate": 1.6292098856804423e-06,
+      "loss": 0.0004,
+      "step": 338
+    },
+    {
+      "clip_ratio": 0.00097316733081243,
+      "epoch": 0.05059465510004198,
+      "grad_norm": 0.04479542747139931,
+      "kl": 0.033306121826171875,
+      "learning_rate": 1.6165959825390661e-06,
+      "loss": 0.0003,
+      "step": 339
+    },
+    {
+      "clip_ratio": 0.001129494600718317,
+      "completion_length": 818.7946815490723,
+      "epoch": 0.05074390187024859,
+      "grad_norm": 0.04621543735265732,
+      "kl": 0.0410919189453125,
+      "learning_rate": 1.604029634760284e-06,
+      "loss": 0.0005,
+      "num_tokens": 17832759.0,
+      "reward": 0.4373923074454069,
+      "reward_std": 0.1885270932689309,
+      "rewards/code_reward": 0.34391014860011637,
+      "rewards/format_reward": 0.9348214343190193,
+      "step": 340
+    },
+    {
+      "clip_ratio": 0.0011062995872634929,
+      "epoch": 0.0508931486404552,
+      "grad_norm": 0.047534842044115067,
+      "kl": 0.042877197265625,
+      "learning_rate": 1.59151136960288e-06,
+      "loss": 0.0004,
+      "step": 341
+    },
+    {
+      "clip_ratio": 0.00121290785864403,
+      "epoch": 0.051042395410661814,
+      "grad_norm": 0.05018023028969765,
+      "kl": 0.04319000244140625,
+      "learning_rate": 1.5790417123081903e-06,
+      "loss": 0.0005,
+      "step": 342
+    },
+    {
+      "clip_ratio": 0.0009951947176887188,
+      "completion_length": 882.1518287658691,
+      "epoch": 0.05119164218086843,
+      "grad_norm": 0.046235859394073486,
+      "kl": 0.03740882873535156,
+      "learning_rate": 1.5666211860780583e-06,
+      "loss": 0.0004,
+      "num_tokens": 18002085.0,
+      "reward": 0.6279045841656625,
+      "reward_std": 0.1975625678896904,
+      "rewards/code_reward": 0.5160295752430102,
+      "rewards/format_reward": 1.1187499947845936,
+      "step": 343
+    },
+    {
+      "clip_ratio": 0.0010207766827079467,
+      "epoch": 0.051340888951075046,
+      "grad_norm": 0.04691191017627716,
+      "kl": 0.037837982177734375,
+      "learning_rate": 1.5542503120528918e-06,
+      "loss": 0.0006,
+      "step": 344
+    },
+    {
+      "clip_ratio": 0.0011089397885371,
+      "epoch": 0.05149013572128166,
+      "grad_norm": 0.047472018748521805,
+      "kl": 0.039302825927734375,
+      "learning_rate": 1.5419296092897866e-06,
+      "loss": 0.0004,
+      "step": 345
+    },
+    {
+      "clip_ratio": 0.00114808726721094,
+      "completion_length": 815.0714664459229,
+      "epoch": 0.05163938249148827,
+      "grad_norm": 0.04975217208266258,
+      "kl": 0.047466278076171875,
+      "learning_rate": 1.529659594740755e-06,
+      "loss": 0.0005,
+      "num_tokens": 18154966.0,
+      "reward": 0.4685569778084755,
+      "reward_std": 0.21453512576408684,
+      "rewards/code_reward": 0.35534268664196134,
+      "rewards/format_reward": 1.1321428623050451,
+      "step": 346
+    },
+    {
+      "clip_ratio": 0.0011890214227605611,
+      "epoch": 0.051788629261694884,
+      "grad_norm": 0.05200071632862091,
+      "kl": 0.04558563232421875,
+      "learning_rate": 1.5174407832310338e-06,
+      "loss": 0.0005,
+      "step": 347
+    },
+    {
+      "clip_ratio": 0.0012163433930254541,
+      "epoch": 0.051937876031901496,
+      "grad_norm": 0.05192115902900696,
+      "kl": 0.0457916259765625,
+      "learning_rate": 1.5052736874374815e-06,
+      "loss": 0.0005,
+      "step": 348
+    },
+    {
+      "clip_ratio": 0.000893519554665545,
+      "completion_length": 747.633960723877,
+      "epoch": 0.05208712280210811,
+      "grad_norm": 0.046751827001571655,
+      "kl": 0.038177490234375,
+      "learning_rate": 1.4931588178670695e-06,
+      "loss": 0.0005,
+      "num_tokens": 18308764.0,
+      "reward": 0.6700454382225871,
+      "reward_std": 0.1688278007786721,
+      "rewards/code_reward": 0.547991847153753,
+      "rewards/format_reward": 1.2205356992781162,
+      "step": 349
+    },
+    {
+      "clip_ratio": 0.00095716763917153,
+      "epoch": 0.05223636957231472,
+      "grad_norm": 0.04715850204229355,
+      "kl": 0.037807464599609375,
+      "learning_rate": 1.4810966828354605e-06,
+      "loss": 0.0005,
+      "step": 350
+    },
+    {
+      "clip_ratio": 0.0010263329295412404,
+      "epoch": 0.05238561634252134,
+      "grad_norm": 0.04827268794178963,
+      "kl": 0.03871917724609375,
+      "learning_rate": 1.469087788445684e-06,
+      "loss": 0.0005,
+      "step": 351
+    },
+    {
+      "clip_ratio": 0.0011772955731430557,
+      "completion_length": 743.0178833007812,
+      "epoch": 0.05253486311272795,
+      "grad_norm": 0.048253390938043594,
+      "kl": 0.0508270263671875,
+      "learning_rate": 1.4571326385668965e-06,
+      "loss": 0.0006,
+      "num_tokens": 18452077.0,
+      "reward": 0.7588480142876506,
+      "reward_std": 0.16709625371731818,
+      "rewards/code_reward": 0.6398301323351916,
+      "rewards/format_reward": 1.1901785582304,
+      "step": 352
+    },
+    {
+      "clip_ratio": 0.0012071774326614104,
+      "epoch": 0.052684109882934566,
+      "grad_norm": 0.04751910641789436,
+      "kl": 0.050937652587890625,
+      "learning_rate": 1.4452317348132434e-06,
+      "loss": 0.0006,
+      "step": 353
+    },
+    {
+      "clip_ratio": 0.001102846614230657,
+      "epoch": 0.05283335665314118,
+      "grad_norm": 0.04808581620454788,
+      "kl": 0.051250457763671875,
+      "learning_rate": 1.4333855765228104e-06,
+      "loss": 0.0006,
+      "step": 354
+    },
+    {
+      "clip_ratio": 0.0012155727981735254,
+      "completion_length": 709.8928928375244,
+      "epoch": 0.05298260342334779,
+      "grad_norm": 0.04265999048948288,
+      "kl": 0.04134368896484375,
+      "learning_rate": 1.421594660736675e-06,
+      "loss": 0.0005,
+      "num_tokens": 18594676.0,
+      "reward": 0.43674828251823783,
+      "reward_std": 0.15885721985250711,
+      "rewards/code_reward": 0.3227304055035347,
+      "rewards/format_reward": 1.140178568661213,
+      "step": 355
+    },
+    {
+      "clip_ratio": 0.0011881247082783375,
+      "epoch": 0.0531318501935544,
+      "grad_norm": 0.055916447192430496,
+      "kl": 0.042873382568359375,
+      "learning_rate": 1.4098594821780476e-06,
+      "loss": 0.0004,
+      "step": 356
+    },
+    {
+      "clip_ratio": 0.0013783813064947026,
+      "epoch": 0.053281096963761015,
+      "grad_norm": 0.0430854894220829,
+      "kl": 0.0421905517578125,
+      "learning_rate": 1.3981805332315174e-06,
+      "loss": 0.0003,
+      "step": 357
+    },
+    {
+      "clip_ratio": 0.0013754095052718185,
+      "completion_length": 803.4911060333252,
+      "epoch": 0.053430343733967635,
+      "grad_norm": 0.05348074436187744,
+      "kl": 0.06780242919921875,
+      "learning_rate": 1.3865583039223929e-06,
+      "loss": 0.0008,
+      "num_tokens": 18753834.0,
+      "reward": 0.5694120684638619,
+      "reward_std": 0.22131631802767515,
+      "rewards/code_reward": 0.4547692039050162,
+      "rewards/format_reward": 1.1464285608381033,
+      "step": 358
+    },
+    {
+      "clip_ratio": 0.0012931995715916855,
+      "epoch": 0.05357959050417425,
+      "grad_norm": 0.16054613888263702,
+      "kl": 0.06353759765625,
+      "learning_rate": 1.374993281896137e-06,
+      "loss": 0.0006,
+      "step": 359
+    },
+    {
+      "clip_ratio": 0.0011385979851183947,
+      "epoch": 0.05372883727438086,
+      "grad_norm": 0.049783170223236084,
+      "kl": 0.0624847412109375,
+      "learning_rate": 1.3634859523979134e-06,
+      "loss": 0.0006,
+      "step": 360
+    },
+    {
+      "clip_ratio": 0.0010799912734000827,
+      "completion_length": 835.8393230438232,
+      "epoch": 0.05387808404458747,
+      "grad_norm": 0.047583021223545074,
+      "kl": 0.037715911865234375,
+      "learning_rate": 1.3520367982522208e-06,
+      "loss": 0.0003,
+      "num_tokens": 18914935.0,
+      "reward": 0.5109263020567596,
+      "reward_std": 0.22399028996005654,
+      "rewards/code_reward": 0.40164059144444764,
+      "rewards/format_reward": 1.0928571410477161,
+      "step": 361
+    },
+    {
+      "clip_ratio": 0.0010377659255027538,
+      "epoch": 0.054027330814794085,
+      "grad_norm": 0.046828076243400574,
+      "kl": 0.037921905517578125,
+      "learning_rate": 1.3406462998426358e-06,
+      "loss": 0.0003,
+      "step": 362
+    },
+    {
+      "clip_ratio": 0.0011691841591527918,
+      "epoch": 0.0541765775850007,
+      "grad_norm": 0.046514738351106644,
+      "kl": 0.0371551513671875,
+      "learning_rate": 1.3293149350916595e-06,
+      "loss": 0.0004,
+      "step": 363
+    },
+    {
+      "clip_ratio": 0.0012077734872946166,
+      "completion_length": 745.857177734375,
+      "epoch": 0.05432582435520731,
+      "grad_norm": 0.046239160001277924,
+      "kl": 0.0434112548828125,
+      "learning_rate": 1.3180431794406623e-06,
+      "loss": 0.0004,
+      "num_tokens": 19068436.0,
+      "reward": 0.6796648311428726,
+      "reward_std": 0.1779340817593038,
+      "rewards/code_reward": 0.5763612519949675,
+      "rewards/format_reward": 1.033035721629858,
+      "step": 364
+    },
+    {
+      "clip_ratio": 0.0012792030256605358,
+      "epoch": 0.05447507112541393,
+      "grad_norm": 0.04772688075900078,
+      "kl": 0.042186737060546875,
+      "learning_rate": 1.3068315058299358e-06,
+      "loss": 0.0004,
+      "step": 365
+    },
+    {
+      "clip_ratio": 0.001383075692501734,
+      "epoch": 0.05462431789562054,
+      "grad_norm": 0.04262588173151016,
+      "kl": 0.04148101806640625,
+      "learning_rate": 1.2956803846788503e-06,
+      "loss": 0.0004,
+      "step": 366
+    },
+    {
+      "clip_ratio": 0.0012452776700229151,
+      "completion_length": 809.1786212921143,
+      "epoch": 0.054773564665827154,
+      "grad_norm": 0.05037975311279297,
+      "kl": 0.044666290283203125,
+      "learning_rate": 1.284590283866116e-06,
+      "loss": 0.0005,
+      "num_tokens": 19220370.0,
+      "reward": 0.3905480457469821,
+      "reward_std": 0.15142776281572878,
+      "rewards/code_reward": 0.2817980256804731,
+      "rewards/format_reward": 1.0874999910593033,
+      "step": 367
+    },
+    {
+      "clip_ratio": 0.0012702639169219765,
+      "epoch": 0.05492281143603377,
+      "grad_norm": 0.04314856976270676,
+      "kl": 0.04655647277832031,
+      "learning_rate": 1.2735616687101518e-06,
+      "loss": 0.0006,
+      "step": 368
+    },
+    {
+      "clip_ratio": 0.0013624547063955106,
+      "epoch": 0.05507205820624038,
+      "grad_norm": 0.05023064464330673,
+      "kl": 0.043704986572265625,
+      "learning_rate": 1.2625950019495614e-06,
+      "loss": 0.0006,
+      "step": 369
+    },
+    {
+      "clip_ratio": 0.0010240589144814294,
+      "completion_length": 683.500020980835,
+      "epoch": 0.05522130497644699,
+      "grad_norm": 0.06126611679792404,
+      "kl": 0.03558349609375,
+      "learning_rate": 1.251690743723718e-06,
+      "loss": 0.0005,
+      "num_tokens": 19360902.0,
+      "reward": 0.8276477949693799,
+      "reward_std": 0.21755714807659388,
+      "rewards/code_reward": 0.7124692033976316,
+      "rewards/format_reward": 1.1517857052385807,
+      "step": 370
+    },
+    {
+      "clip_ratio": 0.0010483075620868476,
+      "epoch": 0.05537055174665361,
+      "grad_norm": 0.07521291077136993,
+      "kl": 0.035892486572265625,
+      "learning_rate": 1.2408493515534581e-06,
+      "loss": 0.0005,
+      "step": 371
+    },
+    {
+      "clip_ratio": 0.0009877505890472094,
+      "epoch": 0.05551979851686022,
+      "grad_norm": 0.061518505215644836,
+      "kl": 0.036228179931640625,
+      "learning_rate": 1.2300712803218834e-06,
+      "loss": 0.0003,
+      "step": 372
+    },
+    {
+      "clip_ratio": 0.0015654707458452322,
+      "completion_length": 786.1607456207275,
+      "epoch": 0.055669045287066836,
+      "grad_norm": 0.04797976464033127,
+      "kl": 0.04172515869140625,
+      "learning_rate": 1.2193569822552772e-06,
+      "loss": 0.0005,
+      "num_tokens": 19512932.0,
+      "reward": 0.6857732241041958,
+      "reward_std": 0.22357454826124012,
+      "rewards/code_reward": 0.5790767734870315,
+      "rewards/format_reward": 1.0669642686843872,
+      "step": 373
+    },
+    {
+      "clip_ratio": 0.0013925148214184446,
+      "epoch": 0.05581829205727345,
+      "grad_norm": 0.047019653022289276,
+      "kl": 0.04467010498046875,
+      "learning_rate": 1.2087069069041268e-06,
+      "loss": 0.0004,
+      "step": 374
+    },
+    {
+      "clip_ratio": 0.0014627039090555627,
+      "epoch": 0.05596753882748006,
+      "grad_norm": 0.04875313118100166,
+      "kl": 0.042537689208984375,
+      "learning_rate": 1.1981215011242654e-06,
+      "loss": 0.0005,
+      "step": 375
+    },
+    {
+      "clip_ratio": 0.0012403775235725334,
+      "completion_length": 799.3482398986816,
+      "epoch": 0.05611678559768667,
+      "grad_norm": 0.04291072115302086,
+      "kl": 0.029933929443359375,
+      "learning_rate": 1.1876012090581184e-06,
+      "loss": 0.0003,
+      "num_tokens": 19668007.0,
+      "reward": 0.4753888142295182,
+      "reward_std": 0.16220126650296152,
+      "rewards/code_reward": 0.3551209402503446,
+      "rewards/format_reward": 1.202678557485342,
+      "step": 376
+    },
+    {
+      "clip_ratio": 0.001163258792985289,
+      "epoch": 0.056266032367893286,
+      "grad_norm": 0.042848821729421616,
+      "kl": 0.031200408935546875,
+      "learning_rate": 1.177146472116071e-06,
+      "loss": 0.0004,
+      "step": 377
+    },
+    {
+      "clip_ratio": 0.0012621104797290172,
+      "epoch": 0.056415279138099905,
+      "grad_norm": 0.04289062321186066,
+      "kl": 0.032794952392578125,
+      "learning_rate": 1.1667577289579462e-06,
+      "loss": 0.0003,
+      "step": 378
+    },
+    {
+      "clip_ratio": 0.0011387938520783791,
+      "completion_length": 887.6250381469727,
+      "epoch": 0.05656452590830652,
+      "grad_norm": 0.04105460271239281,
+      "kl": 0.030487060546875,
+      "learning_rate": 1.1564354154746007e-06,
+      "loss": 0.0003,
+      "num_tokens": 19833942.0,
+      "reward": 0.42751474864780903,
+      "reward_std": 0.2047222899273038,
+      "rewards/code_reward": 0.31001472822390497,
+      "rewards/format_reward": 1.1749999895691872,
+      "step": 379
+    },
+    {
+      "clip_ratio": 0.0010691416100598872,
+      "epoch": 0.05671377267851313,
+      "grad_norm": 0.040559619665145874,
+      "kl": 0.029735565185546875,
+      "learning_rate": 1.146179964769635e-06,
+      "loss": 0.0001,
+      "step": 380
+    },
+    {
+      "clip_ratio": 0.0009864674502750859,
+      "epoch": 0.05686301944871974,
+      "grad_norm": 0.04219922423362732,
+      "kl": 0.030120849609375,
+      "learning_rate": 1.1359918071412195e-06,
+      "loss": 0.0002,
+      "step": 381
+    },
+    {
+      "clip_ratio": 0.001063702917235787,
+      "completion_length": 778.6339664459229,
+      "epoch": 0.057012266218926355,
+      "grad_norm": 0.040541067719459534,
+      "kl": 0.038173675537109375,
+      "learning_rate": 1.1258713700640456e-06,
+      "loss": 0.0004,
+      "num_tokens": 19974552.0,
+      "reward": 0.34642348857596517,
+      "reward_std": 0.1312610807362944,
+      "rewards/code_reward": 0.2391020506620407,
+      "rewards/format_reward": 1.0732142701745033,
+      "step": 382
+    },
+    {
+      "clip_ratio": 0.001328528222074965,
+      "epoch": 0.05716151298913297,
+      "grad_norm": 0.042283330112695694,
+      "kl": 0.038555145263671875,
+      "learning_rate": 1.115819078171383e-06,
+      "loss": 0.0004,
+      "step": 383
+    },
+    {
+      "clip_ratio": 0.001007061529890052,
+      "epoch": 0.05731075975933958,
+      "grad_norm": 0.043051622807979584,
+      "kl": 0.037868499755859375,
+      "learning_rate": 1.1058353532372667e-06,
+      "loss": 0.0004,
+      "step": 384
+    },
+    {
+      "clip_ratio": 0.0009072511438716901,
+      "completion_length": 844.9286155700684,
+      "epoch": 0.0574600065295462,
+      "grad_norm": 0.04384131357073784,
+      "kl": 0.03057098388671875,
+      "learning_rate": 1.0959206141587998e-06,
+      "loss": 0.0003,
+      "num_tokens": 20133619.0,
+      "reward": 0.5458137290552258,
+      "reward_std": 0.11904705758206546,
+      "rewards/code_reward": 0.43795658252201974,
+      "rewards/format_reward": 1.0785714220255613,
+      "step": 385
+    },
+    {
+      "clip_ratio": 0.0010800588406709721,
+      "epoch": 0.05760925329975281,
+      "grad_norm": 0.04235073924064636,
+      "kl": 0.03037261962890625,
+      "learning_rate": 1.0860752769385766e-06,
+      "loss": 0.0002,
+      "step": 386
+    },
+    {
+      "clip_ratio": 0.0009236693222192116,
+      "epoch": 0.057758500069959424,
+      "grad_norm": 0.04210828244686127,
+      "kl": 0.030902862548828125,
+      "learning_rate": 1.0762997546672279e-06,
+      "loss": 0.0001,
+      "step": 387
+    },
+    {
+      "clip_ratio": 0.0011068970561609603,
+      "completion_length": 752.0625419616699,
+      "epoch": 0.05790774684016604,
+      "grad_norm": 0.046190328896045685,
+      "kl": 0.037837982177734375,
+      "learning_rate": 1.0665944575060914e-06,
+      "loss": 0.0006,
+      "num_tokens": 20281186.0,
+      "reward": 0.6457045003771782,
+      "reward_std": 0.2894785313401371,
+      "rewards/code_reward": 0.5300794651557226,
+      "rewards/format_reward": 1.1562499813735485,
+      "step": 388
+    },
+    {
+      "clip_ratio": 0.0010543619564487017,
+      "epoch": 0.05805699361037265,
+      "grad_norm": 0.04631761088967323,
+      "kl": 0.0392608642578125,
+      "learning_rate": 1.056959792669997e-06,
+      "loss": 0.0006,
+      "step": 389
+    },
+    {
+      "clip_ratio": 0.0011255824747422594,
+      "epoch": 0.05820624038057926,
+      "grad_norm": 0.04669871926307678,
+      "kl": 0.038478851318359375,
+      "learning_rate": 1.0473961644101856e-06,
+      "loss": 0.0005,
+      "step": 390
+    },
+    {
+      "clip_ratio": 0.0010092920128954574,
+      "completion_length": 788.1250305175781,
+      "epoch": 0.058355487150785874,
+      "grad_norm": 0.04489421471953392,
+      "kl": 0.03372001647949219,
+      "learning_rate": 1.037903973997345e-06,
+      "loss": 0.0006,
+      "num_tokens": 20439113.0,
+      "reward": 0.7695088880136609,
+      "reward_std": 0.2082495610229671,
+      "rewards/code_reward": 0.6506695952266455,
+      "rewards/format_reward": 1.1883928440511227,
+      "step": 391
+    },
+    {
+      "clip_ratio": 0.0010976425510307308,
+      "epoch": 0.058504733920992494,
+      "grad_norm": 0.04390792176127434,
+      "kl": 0.03353691101074219,
+      "learning_rate": 1.0284836197047737e-06,
+      "loss": 0.0004,
+      "step": 392
+    },
+    {
+      "clip_ratio": 0.0009941325843101367,
+      "epoch": 0.058653980691199106,
+      "grad_norm": 0.041951049119234085,
+      "kl": 0.034313201904296875,
+      "learning_rate": 1.0191354967916712e-06,
+      "loss": 0.0005,
+      "step": 393
+    },
+    {
+      "clip_ratio": 0.0011596001786529087,
+      "completion_length": 790.3571701049805,
+      "epoch": 0.05880322746140572,
+      "grad_norm": 0.043895188719034195,
+      "kl": 0.040134429931640625,
+      "learning_rate": 1.0098599974865515e-06,
+      "loss": 0.0005,
+      "num_tokens": 20595841.0,
+      "reward": 0.529973822645843,
+      "reward_std": 0.24959730240516365,
+      "rewards/code_reward": 0.4122952348552644,
+      "rewards/format_reward": 1.1767856981605291,
+      "step": 394
+    },
+    {
+      "clip_ratio": 0.0011324258230160922,
+      "epoch": 0.05895247423161233,
+      "grad_norm": 0.04367184266448021,
+      "kl": 0.039226531982421875,
+      "learning_rate": 1.0006575109707898e-06,
+      "loss": 0.0004,
+      "step": 395
+    },
+    {
+      "clip_ratio": 0.0011229302381252637,
+      "epoch": 0.059101721001818944,
+      "grad_norm": 0.04425705596804619,
+      "kl": 0.04046630859375,
+      "learning_rate": 9.915284233622877e-07,
+      "loss": 0.0004,
+      "step": 396
+    },
+    {
+      "clip_ratio": 0.0013727044042752823,
+      "completion_length": 719.5178890228271,
+      "epoch": 0.059250967772025556,
+      "grad_norm": 0.04585633426904678,
+      "kl": 0.048259735107421875,
+      "learning_rate": 9.824731176992796e-07,
+      "loss": 0.0005,
+      "num_tokens": 20738538.0,
+      "reward": 0.4819835126399994,
+      "reward_std": 0.1560515824239701,
+      "rewards/code_reward": 0.36680491268998594,
+      "rewards/format_reward": 1.1517857126891613,
+      "step": 397
+    },
+    {
+      "clip_ratio": 0.0012794259218935622,
+      "epoch": 0.05940021454223217,
+      "grad_norm": 0.04646995663642883,
+      "kl": 0.04850006103515625,
+      "learning_rate": 9.734919739242543e-07,
+      "loss": 0.0005,
+      "step": 398
+    },
+    {
+      "clip_ratio": 0.001545508543131291,
+      "epoch": 0.05954946131243879,
+      "grad_norm": 0.04666193947196007,
+      "kl": 0.048038482666015625,
+      "learning_rate": 9.645853688680177e-07,
+      "loss": 0.0005,
+      "step": 399
+    },
+    {
+      "clip_ratio": 0.0013309074929566123,
+      "completion_length": 834.366117477417,
+      "epoch": 0.0596987080826454,
+      "grad_norm": 0.04417085275053978,
+      "kl": 0.03300666809082031,
+      "learning_rate": 9.557536762338786e-07,
+      "loss": 0.0004,
+      "num_tokens": 20898270.0,
+      "reward": 0.4664941602386534,
+      "reward_std": 0.21694895438849926,
+      "rewards/code_reward": 0.342833440809045,
+      "rewards/format_reward": 1.236607126891613,
+      "step": 400
+    },
+    {
+      "clip_ratio": 0.0012776327512256103,
+      "epoch": 0.05984795485285201,
+      "grad_norm": 0.047289684414863586,
+      "kl": 0.03195381164550781,
+      "learning_rate": 9.46997266581973e-07,
+      "loss": 0.0003,
+      "step": 401
+    },
+    {
+      "clip_ratio": 0.001053016601872514,
+      "epoch": 0.059997201623058626,
+      "grad_norm": 0.0457286536693573,
+      "kl": 0.03333282470703125,
+      "learning_rate": 9.383165073137115e-07,
+      "loss": 0.0002,
+      "step": 402
+    },
+    {
+      "clip_ratio": 0.0011895241223101038,
+      "completion_length": 810.7857532501221,
+      "epoch": 0.06014644839326524,
+      "grad_norm": 0.04430186375975609,
+      "kl": 0.035400390625,
+      "learning_rate": 9.297117626563687e-07,
+      "loss": 0.0005,
+      "num_tokens": 21049887.0,
+      "reward": 0.6196182537823915,
+      "reward_std": 0.1633899782318622,
+      "rewards/code_reward": 0.4986360985203646,
+      "rewards/format_reward": 1.2098214030265808,
+      "step": 403
+    },
+    {
+      "clip_ratio": 0.0011378818035154836,
+      "epoch": 0.06029569516347185,
+      "grad_norm": 0.04314730688929558,
+      "kl": 0.03586387634277344,
+      "learning_rate": 9.211833936477957e-07,
+      "loss": 0.0005,
+      "step": 404
+    },
+    {
+      "clip_ratio": 0.001224413937961799,
+      "epoch": 0.06044494193367847,
+      "grad_norm": 0.04237036406993866,
+      "kl": 0.03537178039550781,
+      "learning_rate": 9.127317581212753e-07,
+      "loss": 0.0005,
+      "step": 405
+    },
+    {
+      "clip_ratio": 0.0011261911040492123,
+      "completion_length": 826.6607475280762,
+      "epoch": 0.06059418870388508,
+      "grad_norm": 0.06861373782157898,
+      "kl": 0.0305938720703125,
+      "learning_rate": 9.043572106905084e-07,
+      "loss": 0.0002,
+      "num_tokens": 21215455.0,
+      "reward": 0.694146909750998,
+      "reward_std": 0.2476245230063796,
+      "rewards/code_reward": 0.5715575968497433,
+      "rewards/format_reward": 1.225892849266529,
+      "step": 406
+    },
+    {
+      "clip_ratio": 0.0011172464819537709,
+      "epoch": 0.060743435474091695,
+      "grad_norm": 0.049063630402088165,
+      "kl": 0.0308380126953125,
+      "learning_rate": 8.960601027347321e-07,
+      "loss": 0.0004,
+      "step": 407
+    },
+    {
+      "clip_ratio": 0.0013254524001240497,
+      "epoch": 0.06089268224429831,
+      "grad_norm": 0.04738166183233261,
+      "kl": 0.03113555908203125,
+      "learning_rate": 8.878407823839788e-07,
+      "loss": 0.0004,
+      "step": 408
+    },
+    {
+      "clip_ratio": 0.0012636855608434416,
+      "completion_length": 779.5803909301758,
+      "epoch": 0.06104192901450492,
+      "grad_norm": 0.05109823867678642,
+      "kl": 0.048473358154296875,
+      "learning_rate": 8.796995945044689e-07,
+      "loss": 0.0006,
+      "num_tokens": 21363731.0,
+      "reward": 0.6684678364545107,
+      "reward_std": 0.2355473709758371,
+      "rewards/code_reward": 0.5479321139864624,
+      "rewards/format_reward": 1.205357126891613,
+      "step": 409
+    },
+    {
+      "clip_ratio": 0.0010770953031169483,
+      "epoch": 0.06119117578471153,
+      "grad_norm": 0.05066639184951782,
+      "kl": 0.049713134765625,
+      "learning_rate": 8.716368806841405e-07,
+      "loss": 0.0005,
+      "step": 410
+    },
+    {
+      "clip_ratio": 0.001110375031203148,
+      "epoch": 0.061340422554918145,
+      "grad_norm": 0.04778670892119408,
+      "kl": 0.047908782958984375,
+      "learning_rate": 8.636529792183171e-07,
+      "loss": 0.0004,
+      "step": 411
+    },
+    {
+      "clip_ratio": 0.0012520548352767946,
+      "completion_length": 919.785758972168,
+      "epoch": 0.061489669325124764,
+      "grad_norm": 0.041636690497398376,
+      "kl": 0.036731719970703125,
+      "learning_rate": 8.557482250955144e-07,
+      "loss": 0.0005,
+      "num_tokens": 21526800.0,
+      "reward": 0.4328823662362993,
+      "reward_std": 0.2877360687125474,
+      "rewards/code_reward": 0.32038235204527155,
+      "rewards/format_reward": 1.1250000037252903,
+      "step": 412
+    },
+    {
+      "clip_ratio": 0.001170799987448845,
+      "epoch": 0.06163891609533138,
+      "grad_norm": 0.041270870715379715,
+      "kl": 0.0362396240234375,
+      "learning_rate": 8.479229499833844e-07,
+      "loss": 0.0005,
+      "step": 413
+    },
+    {
+      "clip_ratio": 0.0013163069452275522,
+      "epoch": 0.06178816286553799,
+      "grad_norm": 0.04129687324166298,
+      "kl": 0.036136627197265625,
+      "learning_rate": 8.401774822147976e-07,
+      "loss": 0.0006,
+      "step": 414
+    },
+    {
+      "clip_ratio": 0.0014189317753334763,
+      "completion_length": 816.785758972168,
+      "epoch": 0.0619374096357446,
+      "grad_norm": 0.04516938328742981,
+      "kl": 0.028072357177734375,
+      "learning_rate": 8.325121467740695e-07,
+      "loss": 0.0004,
+      "num_tokens": 21682897.0,
+      "reward": 0.6884303884580731,
+      "reward_std": 0.19168220832943916,
+      "rewards/code_reward": 0.5636089474428445,
+      "rewards/format_reward": 1.2482142746448517,
+      "step": 415
+    },
+    {
+      "clip_ratio": 0.0013228377811174141,
+      "epoch": 0.062086656405951214,
+      "grad_norm": 0.04481777176260948,
+      "kl": 0.027347564697265625,
+      "learning_rate": 8.249272652833226e-07,
+      "loss": 0.0004,
+      "step": 416
+    },
+    {
+      "clip_ratio": 0.0012798961979569867,
+      "epoch": 0.06223590317615783,
+      "grad_norm": 0.045379120856523514,
+      "kl": 0.02756500244140625,
+      "learning_rate": 8.174231559889931e-07,
+      "loss": 0.0004,
+      "step": 417
+    },
+    {
+      "clip_ratio": 0.0010374451630923431,
+      "completion_length": 739.3125247955322,
+      "epoch": 0.06238514994636444,
+      "grad_norm": 0.046955034136772156,
+      "kl": 0.04058074951171875,
+      "learning_rate": 8.100001337484787e-07,
+      "loss": 0.0006,
+      "num_tokens": 21827041.0,
+      "reward": 0.6235413569957018,
+      "reward_std": 0.20960151287727058,
+      "rewards/code_reward": 0.5099699185811915,
+      "rewards/format_reward": 1.135714277625084,
+      "step": 418
+    },
+    {
+      "clip_ratio": 0.0011014225965482183,
+      "epoch": 0.06253439671657106,
+      "grad_norm": 0.04679929092526436,
+      "kl": 0.039306640625,
+      "learning_rate": 8.026585100169251e-07,
+      "loss": 0.0005,
+      "step": 419
+    },
+    {
+      "clip_ratio": 0.0010908065232797526,
+      "epoch": 0.06268364348677767,
+      "grad_norm": 0.047468509525060654,
+      "kl": 0.04001617431640625,
+      "learning_rate": 7.953985928341601e-07,
+      "loss": 0.0005,
+      "step": 420
+    },
+    {
+      "clip_ratio": 0.001171671107840666,
+      "completion_length": 773.3125324249268,
+      "epoch": 0.06283289025698428,
+      "grad_norm": 0.05091605335474014,
+      "kl": 0.042324066162109375,
+      "learning_rate": 7.882206868117693e-07,
+      "loss": 0.0004,
+      "num_tokens": 21981755.0,
+      "reward": 0.6787664843723178,
+      "reward_std": 0.1951366146095097,
+      "rewards/code_reward": 0.5552843157202005,
+      "rewards/format_reward": 1.2348214238882065,
+      "step": 421
+    },
+    {
+      "clip_ratio": 0.000994445233118313,
+      "epoch": 0.0629821370271909,
+      "grad_norm": 0.05059166997671127,
+      "kl": 0.042453765869140625,
+      "learning_rate": 7.81125093120313e-07,
+      "loss": 0.0006,
+      "step": 422
+    },
+    {
+      "clip_ratio": 0.001127321840613149,
+      "epoch": 0.06313138379739751,
+      "grad_norm": 0.05009785294532776,
+      "kl": 0.042308807373046875,
+      "learning_rate": 7.741121094766916e-07,
+      "loss": 0.0006,
+      "step": 423
+    },
+    {
+      "clip_ratio": 0.00141268036350084,
+      "completion_length": 723.9196796417236,
+      "epoch": 0.06328063056760412,
+      "grad_norm": 0.5321958065032959,
+      "kl": 0.40248870849609375,
+      "learning_rate": 7.671820301316532e-07,
+      "loss": 0.0042,
+      "num_tokens": 22127024.0,
+      "reward": 0.6335166869685054,
+      "reward_std": 0.18789594783447683,
+      "rewards/code_reward": 0.5122666736133397,
+      "rewards/format_reward": 1.2124999947845936,
+      "step": 424
+    },
+    {
+      "clip_ratio": 0.0016308118902088609,
+      "epoch": 0.06342987733781073,
+      "grad_norm": 0.3178289532661438,
+      "kl": 0.2588348388671875,
+      "learning_rate": 7.603351458574474e-07,
+      "loss": 0.0027,
+      "step": 425
+    },
+    {
+      "clip_ratio": 0.0017113010744651547,
+      "epoch": 0.06357912410801735,
+      "grad_norm": 0.21873776614665985,
+      "kl": 0.1898651123046875,
+      "learning_rate": 7.535717439356255e-07,
+      "loss": 0.002,
+      "step": 426
+    },
+    {
+      "clip_ratio": 0.001095123085178784,
+      "completion_length": 865.1071853637695,
+      "epoch": 0.06372837087822396,
+      "grad_norm": 0.04840608686208725,
+      "kl": 0.034221649169921875,
+      "learning_rate": 7.46892108144986e-07,
+      "loss": 0.0004,
+      "num_tokens": 22296716.0,
+      "reward": 0.6035498650744557,
+      "reward_std": 0.2631097093690187,
+      "rewards/code_reward": 0.49417484272271395,
+      "rewards/format_reward": 1.0937499850988388,
+      "step": 427
+    },
+    {
+      "clip_ratio": 0.0011084674079029355,
+      "epoch": 0.06387761764843057,
+      "grad_norm": 0.04596693068742752,
+      "kl": 0.034698486328125,
+      "learning_rate": 7.402965187496697e-07,
+      "loss": 0.0004,
+      "step": 428
+    },
+    {
+      "clip_ratio": 0.0010237721726298332,
+      "epoch": 0.0640268644186372,
+      "grad_norm": 0.044551607221364975,
+      "kl": 0.0347747802734375,
+      "learning_rate": 7.337852524873974e-07,
+      "loss": 0.0005,
+      "step": 429
+    },
+    {
+      "clip_ratio": 0.0009593770791980205,
+      "completion_length": 859.4464588165283,
+      "epoch": 0.06417611118884381,
+      "grad_norm": 0.04016838222742081,
+      "kl": 0.0345458984375,
+      "learning_rate": 7.273585825578608e-07,
+      "loss": 0.0005,
+      "num_tokens": 22461077.0,
+      "reward": 0.581423498224467,
+      "reward_std": 0.23203608253970742,
+      "rewards/code_reward": 0.45811990328365937,
+      "rewards/format_reward": 1.2330357059836388,
+      "step": 430
+    },
+    {
+      "clip_ratio": 0.000999740838778962,
+      "epoch": 0.06432535795905042,
+      "grad_norm": 0.03978991135954857,
+      "kl": 0.03399658203125,
+      "learning_rate": 7.21016778611259e-07,
+      "loss": 0.0004,
+      "step": 431
+    },
+    {
+      "clip_ratio": 0.0010453351060277782,
+      "epoch": 0.06447460472925703,
+      "grad_norm": 0.03954523429274559,
+      "kl": 0.034114837646484375,
+      "learning_rate": 7.147601067369835e-07,
+      "loss": 0.0005,
+      "step": 432
+    },
+    {
+      "clip_ratio": 0.0011874185729539022,
+      "completion_length": 807.7143249511719,
+      "epoch": 0.06462385149946365,
+      "grad_norm": 0.04943690821528435,
+      "kl": 0.03006744384765625,
+      "learning_rate": 7.085888294524561e-07,
+      "loss": 0.0003,
+      "num_tokens": 22610992.0,
+      "reward": 0.6454438250511885,
+      "reward_std": 0.2097741151228547,
+      "rewards/code_reward": 0.5236580954515375,
+      "rewards/format_reward": 1.2178571298718452,
+      "step": 433
+    },
+    {
+      "clip_ratio": 0.0009378703916809172,
+      "epoch": 0.06477309826967026,
+      "grad_norm": 0.04660598561167717,
+      "kl": 0.029407501220703125,
+      "learning_rate": 7.025032056921117e-07,
+      "loss": 0.0002,
+      "step": 434
+    },
+    {
+      "clip_ratio": 0.0011401341880628024,
+      "epoch": 0.06492234503987687,
+      "grad_norm": 0.04872843250632286,
+      "kl": 0.02916717529296875,
+      "learning_rate": 6.965034907965349e-07,
+      "loss": 0.0003,
+      "step": 435
+    },
+    {
+      "clip_ratio": 0.0010988110407197382,
+      "completion_length": 820.857177734375,
+      "epoch": 0.06507159181008348,
+      "grad_norm": 0.044269368052482605,
+      "kl": 0.05198097229003906,
+      "learning_rate": 6.905899365017462e-07,
+      "loss": 0.0007,
+      "num_tokens": 22758417.0,
+      "reward": 0.621272899210453,
+      "reward_std": 0.18449996295385063,
+      "rewards/code_reward": 0.5032371644047089,
+      "rewards/format_reward": 1.1803571358323097,
+      "step": 436
+    },
+    {
+      "clip_ratio": 0.001356448690785328,
+      "epoch": 0.0652208385802901,
+      "grad_norm": 0.044534239917993546,
+      "kl": 0.052337646484375,
+      "learning_rate": 6.847627909286409e-07,
+      "loss": 0.0007,
+      "step": 437
+    },
+    {
+      "clip_ratio": 0.0009920232641889015,
+      "epoch": 0.06537008535049671,
+      "grad_norm": 0.05185743048787117,
+      "kl": 0.0524444580078125,
+      "learning_rate": 6.790222985725761e-07,
+      "loss": 0.0004,
+      "step": 438
+    },
+    {
+      "clip_ratio": 0.0011128076639579376,
+      "completion_length": 675.3750305175781,
+      "epoch": 0.06551933212070332,
+      "grad_norm": 0.04222201183438301,
+      "kl": 0.042942047119140625,
+      "learning_rate": 6.733687002931141e-07,
+      "loss": 0.0005,
+      "num_tokens": 22890983.0,
+      "reward": 0.5957542844116688,
+      "reward_std": 0.13940688711591065,
+      "rewards/code_reward": 0.472272110870108,
+      "rewards/format_reward": 1.2348214201629162,
+      "step": 439
+    },
+    {
+      "clip_ratio": 0.0010995728334819432,
+      "epoch": 0.06566857889090993,
+      "grad_norm": 0.0406472273170948,
+      "kl": 0.041461944580078125,
+      "learning_rate": 6.678022333039158e-07,
+      "loss": 0.0005,
+      "step": 440
+    },
+    {
+      "clip_ratio": 0.0010332062774978112,
+      "epoch": 0.06581782566111655,
+      "grad_norm": 0.04296508803963661,
+      "kl": 0.04045867919921875,
+      "learning_rate": 6.623231311627876e-07,
+      "loss": 0.0005,
+      "step": 441
+    },
+    {
+      "clip_ratio": 0.0013383062214415986,
+      "completion_length": 835.9732513427734,
+      "epoch": 0.06596707243132316,
+      "grad_norm": 0.05803956463932991,
+      "kl": 0.047481536865234375,
+      "learning_rate": 6.569316237618811e-07,
+      "loss": 0.0006,
+      "num_tokens": 23046177.0,
+      "reward": 0.6742162802256644,
+      "reward_std": 0.21495943842455745,
+      "rewards/code_reward": 0.5637698317877948,
+      "rewards/format_reward": 1.1044642813503742,
+      "step": 442
+    },
+    {
+      "clip_ratio": 0.0012342822874416015,
+      "epoch": 0.06611631920152979,
+      "grad_norm": 0.04812246188521385,
+      "kl": 0.046344757080078125,
+      "learning_rate": 6.516279373180499e-07,
+      "loss": 0.0003,
+      "step": 443
+    },
+    {
+      "clip_ratio": 0.0012974117453268263,
+      "epoch": 0.0662655659717364,
+      "grad_norm": 0.04777935892343521,
+      "kl": 0.04587554931640625,
+      "learning_rate": 6.464122943633543e-07,
+      "loss": 0.0004,
+      "step": 444
+    },
+    {
+      "clip_ratio": 0.0012328889933996834,
+      "completion_length": 776.0714588165283,
+      "epoch": 0.06641481274194301,
+      "grad_norm": 0.04433201625943184,
+      "kl": 0.03641510009765625,
+      "learning_rate": 6.412849137357271e-07,
+      "loss": 0.0006,
+      "num_tokens": 23188446.0,
+      "reward": 0.4833586225286126,
+      "reward_std": 0.19566731923259795,
+      "rewards/code_reward": 0.36916218511760235,
+      "rewards/format_reward": 1.1419642679393291,
+      "step": 445
+    },
+    {
+      "clip_ratio": 0.0013573511714639608,
+      "epoch": 0.06656405951214962,
+      "grad_norm": 0.04462065547704697,
+      "kl": 0.03658485412597656,
+      "learning_rate": 6.3624601056979e-07,
+      "loss": 0.0005,
+      "step": 446
+    },
+    {
+      "clip_ratio": 0.001113318128773244,
+      "epoch": 0.06671330628235624,
+      "grad_norm": 0.04434194415807724,
+      "kl": 0.03620719909667969,
+      "learning_rate": 6.312957962878278e-07,
+      "loss": 0.0004,
+      "step": 447
+    },
+    {
+      "clip_ratio": 0.0012251644729985856,
+      "completion_length": 870.0357532501221,
+      "epoch": 0.06686255305256285,
+      "grad_norm": 0.04125617817044258,
+      "kl": 0.026210784912109375,
+      "learning_rate": 6.264344785909181e-07,
+      "loss": 0.0004,
+      "num_tokens": 23358459.0,
+      "reward": 0.4485369832254946,
+      "reward_std": 0.12637041695415974,
+      "rewards/code_reward": 0.32514411299780477,
+      "rewards/format_reward": 1.233928557485342,
+      "step": 448
+    },
+    {
+      "clip_ratio": 0.0009692438616184518,
+      "epoch": 0.06701179982276946,
+      "grad_norm": 0.03909595310688019,
+      "kl": 0.026597976684570312,
+      "learning_rate": 6.216622614502149e-07,
+      "loss": 0.0003,
+      "step": 449
+    },
+    {
+      "clip_ratio": 0.001217871102198842,
+      "epoch": 0.06716104659297607,
+      "grad_norm": 0.03862709924578667,
+      "kl": 0.025884628295898438,
+      "learning_rate": 6.169793450983916e-07,
+      "loss": 0.0004,
+      "step": 450
+    },
+    {
+      "clip_ratio": 0.0010045372582681011,
+      "completion_length": 881.0982608795166,
+      "epoch": 0.06731029336318269,
+      "grad_norm": 0.04469547048211098,
+      "kl": 0.0409698486328125,
+      "learning_rate": 6.123859260212393e-07,
+      "loss": 0.0004,
+      "num_tokens": 23516880.0,
+      "reward": 0.5806995194870979,
+      "reward_std": 0.16702931025065482,
+      "rewards/code_reward": 0.4698066448618192,
+      "rewards/format_reward": 1.1089285612106323,
+      "step": 451
+    },
+    {
+      "clip_ratio": 0.0008761170529396622,
+      "epoch": 0.0674595401333893,
+      "grad_norm": 0.04609547555446625,
+      "kl": 0.04105377197265625,
+      "learning_rate": 6.07882196949423e-07,
+      "loss": 0.0004,
+      "step": 452
+    },
+    {
+      "clip_ratio": 0.0009610654269636143,
+      "epoch": 0.06760878690359591,
+      "grad_norm": 0.04481823369860649,
+      "kl": 0.040618896484375,
+      "learning_rate": 6.034683468503948e-07,
+      "loss": 0.0004,
+      "step": 453
+    },
+    {
+      "clip_ratio": 0.000957371406002494,
+      "completion_length": 678.1160984039307,
+      "epoch": 0.06775803367380252,
+      "grad_norm": 0.046332020312547684,
+      "kl": 0.0386810302734375,
+      "learning_rate": 5.991445609204641e-07,
+      "loss": 0.0004,
+      "num_tokens": 23650313.0,
+      "reward": 0.7209968403913081,
+      "reward_std": 0.17456804146058857,
+      "rewards/code_reward": 0.5993896879081149,
+      "rewards/format_reward": 1.216071404516697,
+      "step": 454
+    },
+    {
+      "clip_ratio": 0.000984678908935166,
+      "epoch": 0.06790728044400914,
+      "grad_norm": 0.04585450142621994,
+      "kl": 0.0377960205078125,
+      "learning_rate": 5.949110205770292e-07,
+      "loss": 0.0003,
+      "step": 455
+    },
+    {
+      "clip_ratio": 0.0011189606830157572,
+      "epoch": 0.06805652721421576,
+      "grad_norm": 0.07117211818695068,
+      "kl": 0.037078857421875,
+      "learning_rate": 5.90767903450964e-07,
+      "loss": 0.0004,
+      "step": 456
+    },
+    {
+      "clip_ratio": 0.0012076381026417948,
+      "completion_length": 843.410758972168,
+      "epoch": 0.06820577398442237,
+      "grad_norm": 0.043919820338487625,
+      "kl": 0.03281402587890625,
+      "learning_rate": 5.867153833791652e-07,
+      "loss": 0.0004,
+      "num_tokens": 23817435.0,
+      "reward": 0.7192922285757959,
+      "reward_std": 0.18404422770254314,
+      "rewards/code_reward": 0.5991136515513062,
+      "rewards/format_reward": 1.2017856948077679,
+      "step": 457
+    },
+    {
+      "clip_ratio": 0.0012479067445383407,
+      "epoch": 0.06835502075462899,
+      "grad_norm": 0.044375985860824585,
+      "kl": 0.032871246337890625,
+      "learning_rate": 5.827536303972587e-07,
+      "loss": 0.0003,
+      "step": 458
+    },
+    {
+      "clip_ratio": 0.0012894814190076431,
+      "epoch": 0.0685042675248356,
+      "grad_norm": 0.045444510877132416,
+      "kl": 0.032073974609375,
+      "learning_rate": 5.78882810732465e-07,
+      "loss": 0.0004,
+      "step": 459
+    },
+    {
+      "clip_ratio": 0.0011302930452075088,
+      "completion_length": 826.9107437133789,
+      "epoch": 0.06865351429504221,
+      "grad_norm": 0.046317800879478455,
+      "kl": 0.033939361572265625,
+      "learning_rate": 5.75103086796625e-07,
+      "loss": 0.0005,
+      "num_tokens": 23968541.0,
+      "reward": 0.5424030870199203,
+      "reward_std": 0.21855923091061413,
+      "rewards/code_reward": 0.4215994896367192,
+      "rewards/format_reward": 1.208035696297884,
+      "step": 460
+    },
+    {
+      "clip_ratio": 0.001294198547839187,
+      "epoch": 0.06880276106524882,
+      "grad_norm": 0.043790195137262344,
+      "kl": 0.03406524658203125,
+      "learning_rate": 5.714146171793846e-07,
+      "loss": 0.0005,
+      "step": 461
+    },
+    {
+      "clip_ratio": 0.0012374374346109107,
+      "epoch": 0.06895200783545544,
+      "grad_norm": 0.07143299281597137,
+      "kl": 0.0339813232421875,
+      "learning_rate": 5.678175566415422e-07,
+      "loss": 0.0004,
+      "step": 462
+    },
+    {
+      "clip_ratio": 0.0014906306660122937,
+      "completion_length": 904.3839683532715,
+      "epoch": 0.06910125460566205,
+      "grad_norm": 0.051049333065748215,
+      "kl": 0.024951934814453125,
+      "learning_rate": 5.643120561085528e-07,
+      "loss": 0.0003,
+      "num_tokens": 24133308.0,
+      "reward": 0.5312421112321317,
+      "reward_std": 0.27633541519753635,
+      "rewards/code_reward": 0.41061710997018963,
+      "rewards/format_reward": 1.2062499895691872,
+      "step": 463
+    },
+    {
+      "clip_ratio": 0.0014828559123998275,
+      "epoch": 0.06925050137586866,
+      "grad_norm": 0.04627595841884613,
+      "kl": 0.02480316162109375,
+      "learning_rate": 5.608982626641991e-07,
+      "loss": 0.0003,
+      "step": 464
+    },
+    {
+      "clip_ratio": 0.0012723907530016731,
+      "epoch": 0.06939974814607527,
+      "grad_norm": 0.0477118119597435,
+      "kl": 0.025279998779296875,
+      "learning_rate": 5.575763195444166e-07,
+      "loss": 0.0003,
+      "step": 465
+    },
+    {
+      "clip_ratio": 0.0012686359677900327,
+      "completion_length": 834.9107551574707,
+      "epoch": 0.06954899491628189,
+      "grad_norm": 0.04587775841355324,
+      "kl": 0.0385589599609375,
+      "learning_rate": 5.543463661312847e-07,
+      "loss": 0.0005,
+      "num_tokens": 24281201.0,
+      "reward": 0.6464886767789721,
+      "reward_std": 0.1595687207300216,
+      "rewards/code_reward": 0.533274372224696,
+      "rewards/format_reward": 1.132142849266529,
+      "step": 466
+    },
+    {
+      "clip_ratio": 0.0011379432999092387,
+      "epoch": 0.0696982416864885,
+      "grad_norm": 0.04559098929166794,
+      "kl": 0.038360595703125,
+      "learning_rate": 5.512085379471808e-07,
+      "loss": 0.0005,
+      "step": 467
+    },
+    {
+      "clip_ratio": 0.0011667834351101192,
+      "epoch": 0.06984748845669511,
+      "grad_norm": 0.045429207384586334,
+      "kl": 0.03958892822265625,
+      "learning_rate": 5.481629666490903e-07,
+      "loss": 0.0005,
+      "step": 468
+    },
+    {
+      "clip_ratio": 0.001148404139712511,
+      "completion_length": 808.4821834564209,
+      "epoch": 0.06999673522690172,
+      "grad_norm": 0.04807772859930992,
+      "kl": 0.03765106201171875,
+      "learning_rate": 5.452097800230853e-07,
+      "loss": 0.0005,
+      "num_tokens": 24434492.0,
+      "reward": 0.4852910442277789,
+      "reward_std": 0.24017929495312274,
+      "rewards/code_reward": 0.3615410323254764,
+      "rewards/format_reward": 1.2374999970197678,
+      "step": 469
+    },
+    {
+      "clip_ratio": 0.001201979286634014,
+      "epoch": 0.07014598199710835,
+      "grad_norm": 0.04833844304084778,
+      "kl": 0.035778045654296875,
+      "learning_rate": 5.423491019789623e-07,
+      "loss": 0.0005,
+      "step": 470
+    },
+    {
+      "clip_ratio": 0.0010972645086440025,
+      "epoch": 0.07029522876731496,
+      "grad_norm": 0.04807697981595993,
+      "kl": 0.035228729248046875,
+      "learning_rate": 5.395810525450425e-07,
+      "loss": 0.0005,
+      "step": 471
+    },
+    {
+      "clip_ratio": 0.0011596989388635848,
+      "completion_length": 822.7946758270264,
+      "epoch": 0.07044447553752158,
+      "grad_norm": 0.047449301928281784,
+      "kl": 0.033260345458984375,
+      "learning_rate": 5.369057478631359e-07,
+      "loss": 0.0003,
+      "num_tokens": 24599963.0,
+      "reward": 0.49419981660321355,
+      "reward_std": 0.2545440453104675,
+      "rewards/code_reward": 0.38009265484288335,
+      "rewards/format_reward": 1.1410714071244001,
+      "step": 472
+    },
+    {
+      "clip_ratio": 0.001089409208361758,
+      "epoch": 0.07059372230772819,
+      "grad_norm": 0.045311130583286285,
+      "kl": 0.03286552429199219,
+      "learning_rate": 5.343233001836694e-07,
+      "loss": 0.0003,
+      "step": 473
+    },
+    {
+      "clip_ratio": 0.0011317514581605792,
+      "epoch": 0.0707429690779348,
+      "grad_norm": 0.050007469952106476,
+      "kl": 0.03356742858886719,
+      "learning_rate": 5.318338178609754e-07,
+      "loss": 0.0003,
+      "step": 474
+    },
+    {
+      "clip_ratio": 0.0010420075359434122,
+      "completion_length": 740.9911155700684,
+      "epoch": 0.07089221584814141,
+      "grad_norm": 0.04966285079717636,
+      "kl": 0.045757293701171875,
+      "learning_rate": 5.294374053487459e-07,
+      "loss": 0.0005,
+      "num_tokens": 24747529.0,
+      "reward": 0.5183156300336123,
+      "reward_std": 0.20554508129134774,
+      "rewards/code_reward": 0.40831562504172325,
+      "rewards/format_reward": 1.0999999940395355,
+      "step": 475
+    },
+    {
+      "clip_ratio": 0.001186353631055681,
+      "epoch": 0.07104146261834803,
+      "grad_norm": 0.04935722425580025,
+      "kl": 0.046184539794921875,
+      "learning_rate": 5.271341631956511e-07,
+      "loss": 0.0006,
+      "step": 476
+    },
+    {
+      "clip_ratio": 0.001091830434233998,
+      "epoch": 0.07119070938855464,
+      "grad_norm": 0.04833516106009483,
+      "kl": 0.04679107666015625,
+      "learning_rate": 5.249241880411181e-07,
+      "loss": 0.0005,
+      "step": 477
+    },
+    {
+      "clip_ratio": 0.0012747206210406148,
+      "completion_length": 907.5357551574707,
+      "epoch": 0.07133995615876125,
+      "grad_norm": 0.04407478868961334,
+      "kl": 0.029878616333007812,
+      "learning_rate": 5.228075726112785e-07,
+      "loss": 0.0003,
+      "num_tokens": 24916653.0,
+      "reward": 0.38216201262548566,
+      "reward_std": 0.2679919386282563,
+      "rewards/code_reward": 0.2559120004007127,
+      "rewards/format_reward": 1.2624999769032001,
+      "step": 478
+    },
+    {
+      "clip_ratio": 0.0011643571451713797,
+      "epoch": 0.07148920292896786,
+      "grad_norm": 0.044017162173986435,
+      "kl": 0.029003143310546875,
+      "learning_rate": 5.207844057150768e-07,
+      "loss": 0.0002,
+      "step": 479
+    },
+    {
+      "clip_ratio": 0.0011917059473489644,
+      "epoch": 0.07163844969917448,
+      "grad_norm": 0.04939346760511398,
+      "kl": 0.029766082763671875,
+      "learning_rate": 5.188547722405437e-07,
+      "loss": 0.0002,
+      "step": 480
+    },
+    {
+      "clip_ratio": 0.0012131860949011752,
+      "completion_length": 848.0000305175781,
+      "epoch": 0.07178769646938109,
+      "grad_norm": 0.04608817771077156,
+      "kl": 0.0340423583984375,
+      "learning_rate": 5.170187531512351e-07,
+      "loss": 0.0004,
+      "num_tokens": 25079536.0,
+      "reward": 0.5753255924209952,
+      "reward_std": 0.2258681992534548,
+      "rewards/code_reward": 0.4602362932055257,
+      "rewards/format_reward": 1.1508928388357162,
+      "step": 481
+    },
+    {
+      "clip_ratio": 0.0014426689594984055,
+      "epoch": 0.0719369432395877,
+      "grad_norm": 0.04564032703638077,
+      "kl": 0.03497314453125,
+      "learning_rate": 5.152764254828348e-07,
+      "loss": 0.0005,
+      "step": 482
+    },
+    {
+      "clip_ratio": 0.0012913148202642333,
+      "epoch": 0.07208619000979431,
+      "grad_norm": 0.04598577320575714,
+      "kl": 0.034259796142578125,
+      "learning_rate": 5.136278623399225e-07,
+      "loss": 0.0004,
+      "step": 483
+    },
+    {
+      "clip_ratio": 0.0010959333521896042,
+      "completion_length": 809.1339664459229,
+      "epoch": 0.07223543678000094,
+      "grad_norm": 0.04507589712738991,
+      "kl": 0.028133392333984375,
+      "learning_rate": 5.120731328929058e-07,
+      "loss": 0.0002,
+      "num_tokens": 25236876.0,
+      "reward": 0.5080648576840758,
+      "reward_std": 0.2463505994528532,
+      "rewards/code_reward": 0.38181484618689865,
+      "rewards/format_reward": 1.2624999918043613,
+      "step": 484
+    },
+    {
+      "clip_ratio": 0.0009193003133987077,
+      "epoch": 0.07238468355020755,
+      "grad_norm": 0.048073090612888336,
+      "kl": 0.029356002807617188,
+      "learning_rate": 5.106123023751187e-07,
+      "loss": 0.0002,
+      "step": 485
+    },
+    {
+      "clip_ratio": 0.0010504390702408273,
+      "epoch": 0.07253393032041416,
+      "grad_norm": 0.07422798126935959,
+      "kl": 0.028322219848632812,
+      "learning_rate": 5.092454320800833e-07,
+      "loss": 0.0002,
+      "step": 486
+    },
+    {
+      "clip_ratio": 0.0009964609798771562,
+      "completion_length": 852.1696815490723,
+      "epoch": 0.07268317709062078,
+      "grad_norm": 0.04429541900753975,
+      "kl": 0.030139923095703125,
+      "learning_rate": 5.079725793589405e-07,
+      "loss": 0.0004,
+      "num_tokens": 25399218.0,
+      "reward": 0.631489341147244,
+      "reward_std": 0.18184524704702199,
+      "rewards/code_reward": 0.5168464637827128,
+      "rewards/format_reward": 1.1464285738766193,
+      "step": 487
+    },
+    {
+      "clip_ratio": 0.0010598115204629721,
+      "epoch": 0.07283242386082739,
+      "grad_norm": 0.04628746956586838,
+      "kl": 0.030010223388671875,
+      "learning_rate": 5.067937976180407e-07,
+      "loss": 0.0005,
+      "step": 488
+    },
+    {
+      "clip_ratio": 0.000984342819720041,
+      "epoch": 0.072981670631034,
+      "grad_norm": 0.044439591467380524,
+      "kl": 0.030406951904296875,
+      "learning_rate": 5.057091363167046e-07,
+      "loss": 0.0002,
+      "step": 489
+    },
+    {
+      "clip_ratio": 0.001186176219562185,
+      "completion_length": 868.383960723877,
+      "epoch": 0.07313091740124061,
+      "grad_norm": 0.05477144569158554,
+      "kl": 0.03466987609863281,
+      "learning_rate": 5.047186409651489e-07,
+      "loss": 0.0005,
+      "num_tokens": 25559141.0,
+      "reward": 0.543341277167201,
+      "reward_std": 0.215598865179345,
+      "rewards/code_reward": 0.428251969628036,
+      "rewards/format_reward": 1.150892861187458,
+      "step": 490
+    },
+    {
+      "clip_ratio": 0.0009668781112850411,
+      "epoch": 0.07328016417144723,
+      "grad_norm": 0.04920962452888489,
+      "kl": 0.035614013671875,
+      "learning_rate": 5.038223531225742e-07,
+      "loss": 0.0005,
+      "step": 491
+    },
+    {
+      "clip_ratio": 0.001169070330888644,
+      "epoch": 0.07342941094165384,
+      "grad_norm": 0.05034809932112694,
+      "kl": 0.035518646240234375,
+      "learning_rate": 5.030203103954232e-07,
+      "loss": 0.0005,
+      "step": 492
+    },
+    {
+      "clip_ratio": 0.0012396883666951908,
+      "completion_length": 859.9464721679688,
+      "epoch": 0.07357865771186045,
+      "grad_norm": 0.044976621866226196,
+      "kl": 0.029693603515625,
+      "learning_rate": 5.023125464358026e-07,
+      "loss": 0.0002,
+      "num_tokens": 25719918.0,
+      "reward": 0.27847930788993835,
+      "reward_std": 0.13773812353610992,
+      "rewards/code_reward": 0.15937215336816735,
+      "rewards/format_reward": 1.1910714134573936,
+      "step": 493
+    },
+    {
+      "clip_ratio": 0.001113424419600051,
+      "epoch": 0.07372790448206706,
+      "grad_norm": 0.044692691415548325,
+      "kl": 0.0286712646484375,
+      "learning_rate": 5.016990909400709e-07,
+      "loss": 0.0002,
+      "step": 494
+    },
+    {
+      "clip_ratio": 0.0011506410692163627,
+      "epoch": 0.07387715125227368,
+      "grad_norm": 0.043276555836200714,
+      "kl": 0.02989959716796875,
+      "learning_rate": 5.011799696475915e-07,
+      "loss": 0.0003,
+      "step": 495
+    },
+    {
+      "clip_ratio": 0.0011127546786156017,
+      "completion_length": 859.4643325805664,
+      "epoch": 0.07402639802248029,
+      "grad_norm": 0.04166851192712784,
+      "kl": 0.037921905517578125,
+      "learning_rate": 5.007552043396547e-07,
+      "loss": 0.0005,
+      "num_tokens": 25887123.0,
+      "reward": 0.5738976923748851,
+      "reward_std": 0.21996640338329598,
+      "rewards/code_reward": 0.4523798130394425,
+      "rewards/format_reward": 1.2151785753667355,
+      "step": 496
+    },
+    {
+      "clip_ratio": 0.0009825457909755642,
+      "epoch": 0.07417564479268692,
+      "grad_norm": 0.04101413115859032,
+      "kl": 0.036708831787109375,
+      "learning_rate": 5.004248128385618e-07,
+      "loss": 0.0004,
+      "step": 497
+    },
+    {
+      "clip_ratio": 0.0010075290774693713,
+      "epoch": 0.07432489156289353,
+      "grad_norm": 0.04146130010485649,
+      "kl": 0.03801918029785156,
+      "learning_rate": 5.001888090068784e-07,
+      "loss": 0.0005,
+      "step": 498
+    },
+    {
+      "clip_ratio": 0.0013720846254727803,
+      "completion_length": 875.857177734375,
+      "epoch": 0.07447413833310014,
+      "grad_norm": 0.05116493999958038,
+      "kl": 0.0377960205078125,
+      "learning_rate": 5.000472027468528e-07,
+      "loss": 0.0004,
+      "num_tokens": 26055779.0,
+      "reward": 0.5694272452965379,
+      "reward_std": 0.2635777585674077,
+      "rewards/code_reward": 0.45674864642933244,
+      "rewards/format_reward": 1.1267857141792774,
+      "step": 499
+    },
+    {
+      "clip_ratio": 0.0013736705577684916,
+      "epoch": 0.07462338510330675,
+      "grad_norm": 0.055545829236507416,
+      "kl": 0.038105010986328125,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 0.0005,
+      "step": 500
+    },
+    {
+      "epoch": 0.07462338510330675,
+      "step": 500,
+      "total_flos": 0.0,
+      "train_loss": 0.0003596346975994038,
+      "train_runtime": 186017.9741,
+      "train_samples_per_second": 0.301,
+      "train_steps_per_second": 0.003
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 500,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}